Influence of Tactile Feedback on Trunk Coordination and  
Electromyography During Walking 
 
by 
 
Braden Heath Romer 
 
 
 
A dissertation submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Doctor of Philosophy 
 
Auburn, Alabama 
December 14, 2013 
 
 
 
 
Keywords: gait coordination, thorax-pelvis relative phase, shod gait, barefoot gait, kinematics, 
walking, footwear, biomechanics, electromyography 
 
 
Copyright 2013 by Braden Heath Romer 
 
 
Approved by 
 
Wendi H. Weimar, Chair, Associate Professor of Kinesiology 
Sean Gallagher, Associate Professor of Industrial & Systems Engineering 
Nels H. Madsen, Professor of Mechanical Engineering 
Matthew W. Miller, Assistant Professor of Kinesiology 
Gretchen D. Oliver, Assistant Professor of Kinesiology 
ii 
 
Abstract 
 Footwear is a necessary component of modern life, serving wide-ranging purposes from 
fashion apparel to personal protective equipment. Despite what would appear as an otherwise 
obscure addition of clothing or equipment for individuals, past research has indicated 
considerable footwear effects on spatiotemporal, kinematics, and kinetics during walking and 
running. Likewise, footwear has been implicated in altered lumbar and pelvic positioning while 
standing. While early gait-related research focused on descriptive information to examine the 
relative timing of body segments throughout the gait cycle, the relationship between trunk, 
pelvic, and lower extremity movements in maintaining forward momentum and increasing 
economy during gait is not yet fully understood. Furthermore, despite the considerable footwear 
effects on gait mechanics, it is not yet known whether footwear has a significant effect on 
segment synchronicity during walking. Therefore, the purposes of this investigation was: (1) to 
investigate the effects of velocity and cadence on lumbopelvic rhythm during overground 
walking; (2) to investigate the effects of footwear on lumbopelvic rhythm during gait; (3) to 
examine the effects of enhanced tactile feedback (i.e. barefoot, textured insole) on the onset of 
muscle activity at foot strike; (4) to examine the effects of spatiotemporal and footwear variation 
on the magnitude of lumbar (i.e. erector spinae) and lower extremity muscular activity (e.g. 
gluteus medius, biceps femoris, soleus, peroneus longus) during the stance phase of gait, and (5) 
to investigate differences in the kinematics and electromyography between males and females.
iii 
 
 Results indicated significant increases in height normalized stride length, cadence, and 
velocity as spatiotemporal variables were manipulated from 75% of Normal, to Normal, and 
125% of Normal. Results also indicated that altering velocity or cadence had a significant effect 
for TP mean, TP SD, TL mean, TL SD, PL mean, PL SD, Pelvic Rotation, Trunk Rotation, and Leg 
Rotation; however the influence of velocity appeared to be smaller in which significant 
differences were found only for TP mean, Pelvic Rotation, and Trunk Rotation. Likewise, 
normalized peak sEMG during the stance phase increased as velocity and cadence increased 
from 75% of Normal, to Normal, to 125% of Normal for lower extremity and lumbar 
musculature. 
When examining footwear effects, results indicated that footwear had a significant effect 
on the preferred walking velocity, cadence, and normalize stride length of the participants. 
Specifically, the barefoot and barefoot-like condition (insole-only) exhibited a significantly 
shorter normalized stride length and lower velocity relative to both shod conditions (shod and 
insole-shod). Conversely, the present study found no significant footwear effect on normalized 
peak muscle activity during the stance phase or onset of muscle activity at foot strike. The only 
kinematic variable to display a significant footwear effect was the relative phase of the pelvis-
leg, with barefoot and insole-only exhibiting greater asynchronicity than the shod and insole-
shod conditions. 
 Finally, the present study revealed some gender differences are present during walking. 
For instance, females exhibited a significantly longer normalized stride length and higher 
cadence than males for the altered footwear conditions only. Conversely, there were no 
significant differences between males and females for during any of the altered walking velocity 
or cadence conditions. Relative phase differences were only found to be significantly different 
iv 
 
during the footwear conditions, with males exhibiting a significantly greater TP SD, indicating 
that the trunk-pelvis of males was significantly more variable than females when footwear was 
altered. Finally, no significant gender differences were noted for normalized peak muscle activity 
during the stance phase of walking during altered velocity, altered cadence, or altered footwear. 
v 
 
Acknowledgments 
The author would like to thank the members of his committee, Dr. Wendi Weimar, Dr. 
Sean Gallagher, Dr. Nels Madsen, Dr. Matthew Miller, and Dr. Gretchen Oliver for their time, 
patience, and dedication to his academic development and career at Auburn University. Their 
careful guidance and expertise was instrumental to the success of the present project. The author 
would also like to thank Dr. Robert Thomas for his willingness to serve as the outside reader.  
 The author is appreciative of the in-kind product donation of Adidas and Algeos Limited. 
Without the considerable product support of these two companies, the present project would 
never have even been started.  
The author also thanks the members of the Sport Biomechanics Laboratory: John Fox, 
Jared Rehm, and Adam Jagodinsky who spent many hours, days, and even holidays aiding in the 
completion of the project. Mr. John Fox and Ms. Taylor Buchanan are greatly thanked for their 
instrumental support and help during the data reduction process. 
 While there is no possible way to adequately convey his thanks to Dr. Wendi Weimar, 
the author sincerely appreciates the significant amount of time she has invested in his 
development over the last several years. Dr. Weimar?s ongoing mentorship has pushed the author 
to yearn for higher standards as an educator, researcher, and as a person. Without her significant 
effort and belief in the abilities of the author, the initiation and completion of the present project 
would never have been possible. 
 Finally, the author is well aware that the path to the completion of the present project 
could have prematurely ended at many points without the considerable support of his family. 
vi 
 
He would like to thank his parents, Keith A. Romer and Rebecca Woods, as well his step-father, 
Thomas Woods, and grandparents, Charlotte McCulty, Don Romer, and Dr. Nidia Romer. The 
author?s brothers, Keith D. Romer and Shawn McCulty, have provided a constant source of 
inspiration and motivation. The author is appreciative of the many direct and indirect lessons that 
they have taught and shared with him throughout the years. It is in the memory of his brother 
Shawn that the author dedicates the completion of this project.  
Perhaps most importantly though, the author would like to thank the one individual that 
has consistently motivated him to push beyond his preconceived notions of limitations, his best 
friend and wife, Terrieha Romer. Undoubtedly, without her wisdom, support, and love, the 
author could have never imagined the possibility of such an amazing journey. 
 
 
 
 
 
 
 
 
 
 
 
 
vii 
 
Table of Contents 
Abstract?????????????????????????????????........ii
Acknowledgments???????????????????????????????v  
List of Tables????????????????????????????????..ix  
List of Figures????????????????????????????????xi  
Chapter I. Introduction?????????????????????????????.1  
 Purpose???????????????????????????????...10  
 Hypotheses??????????????????????????????.11  
 Limitations??????????????????????????????.11  
 Delimitations?????????????????????????????..11  
 Definition of Terms???????????????????????????12  
Chapter II: Review of Literature?????????????????????????14  
 Section 1: Human Gait and the Phases of Walking??????????????...15  
 Section 2: Role of the Lumbopelvic Region on Human Locomotion???????...2 7 
 Section 3: Influence of Footwear on Human Gait??????????????? .38 
Section 4: Role of Plantar Sensation and Tactile Feedback in Human Locomotion??.44  
Section 5: Summary of the Literature???????????????????? 49 
Chapter III: Methodology???????????????????????????..52  
 Participants??????????????????????????????53  
 Setting????????????????????????????????54
viii 
 
 Instrumentation????????????????????????????..54  
 Design & Procedures??????????????????????????.62  
Experimental Design & Data Analysis???????????????????..65  
Section IV: Results????????????????????????????...?.70  
Section 1: Participant Demographics????????????????????.71  
Section 2: Walking Velocity Effects????????????????????.72  
Section 3: Cadence Effects???????????????????????....89  
 Section 4: Footwear Effects???????????????????????.107  
Chapter V: Discussion????????????????????????????.116  
 Section 1: Walking Velocity Effects????????????????????116  
 Section 2: Walking Cadence Effects????????????????????121  
 Section 3: Footwear Effects during Walking????????????????...128  
 Section 4: Summary of the Discussion? ??????????????????1 34 
 Section 5: Future Research???????????????????????..1 43 
References????????????????????????????????....1 45 
Appendix A: Health History Questionnaire???????????????????? .155 
Appendix B: Informed Consent????????????????????????.. .156 
 
 
ix 
 
List of Tables 
Table 1: Percentages of Stance and Swing Phases during Gait?????????????.18  
Table 2: Names and Positions of Markers Used in Modified Plug-In-Gait Model..?????.56  
Table 3: Participant Demographics???????????????????????....72  
Table 4: Kinematic Variables: Velocity Effects??????????????????? 73 
Table 5: Non-Significant Condition Effects for Kinematics when Velocity was Varied???..80  
Table 6: sEMG Variables??????????????????????????? ...81 
Table 7: Non-Significant Condition Effects for sEMG When Velocity was Varied?????.87  
Table 8: Kinematic Variables: Cadence Effects???????????????????89  
Table 9: Non-Significant Condition Effects for sEMG When Cadence was Varied???? ...105 
Table 10: Kinematic Variables: Footwear Effects?????????????????... 107 
Table 11: Non-Significant Condition Effects for Kinematics When Footwear was Varied??11 2 
Table 12: Non-significant Gender Effects when Footwear was Varied?????????..11 4 
Table 13: sEMG Variables: Footwear Effects?????????????? ???? .?11 5 
Table 14: Effect of Velocity on all Kinematic Variables as Compared to Normal?????117  
Table 15: Effect of Velocity on all sEMG Variables as Compared to Normal??????...118  
Table 16: Effect of Cadence on all Kinematic Variables as Compared to Normal?????.122  
Table 17: Effect of Cadence on all sEMG Variables as Compared to Normal??????...123  
Table 18: Changes in Normalized Stride Length during Altered Velocity & Altered Cadence.125 
Table 19: Changes in Velocity during Altered Velocity & Altered Cadence???????.125  
Table 20: Changes in Cadence during Altered Velocity & Altered Cadence???????.126  
x 
 
Table 21: Effect of Footwear on all Kinematic Variables as Compared to Shod?????...1 29 
Table 22: Effect of Footwear on all sEMG Variables as Compared to Shod???????.1 31
xii 
 
List of Figures 
Figure 1: Phases of Gait???????????????????????????? .16 
Figure 2: Functional Phases of Gait??..?????????????????????17  
Figure 3: Gait Cycle Divisions????..?????????????????????20  
Figure 4: Six Determinants of Gait: Pelvic Rotation????????????????.?25  
Figure 5: Six Determinants of Gait: Contralateral Drop????????????????25  
Figure 6: Six Determinants of Gait: Lateral Displacement??????????????...2 6 
Figure 7: Six Determinants of Gait: Foot & Ankle Motion?? ????????????...27  
Figure 8: Six Determinants of Gait: Knee Motion???..??????????????...27  
Figure 9: Vicon Motion Capture Camera Orientation????????????????...55  
Figure 10: GAITRite Mat??????????...????????????????... 57 
Figure 11: Noraxon Telemyo 2400T V2 Transmitter?????????????????58  
Figure 12: Noraxon Telemyo 2400R World Wide Telemetry Receiver??????????58  
Figure 13: Footwear Utilized in the Present Study?????????????????? 61 
Figure 14: Textured Insole Utilized in the Present Study?????????????? ?..62  
Figure 15: Huarache Sandal Design Utilized in the Present Study for Insole Only Condition? .62 
Figure 16: Effect of Walking Velocity on Cadence?????????????????.. .74 
Figure 17: Effects of Walking Velocity on Normalized Stride Length?????????? .75 
Figure 18: Example Phase Portrait of the Pelvis during One-Stride???????????.76  
Figure 19: Effect of Walking Velocity on TP mean??????????????????.77  
Figure 20: Effect of Walking Velocity on Pelvic Rotation?????????????. ..?78  
Figure 21: Effect of Walking Velocity on Trunk Rotation?????????????...?79  
Figure 22: Effect of Walking Velocity on PerL peak during Stance???????????.. .82 
xiii 
 
Figure 23: Effect of Walking Velocity on SOL Peak during Stance??????????..? .83 
Figure 24: Effect of Walking Velocity on BF Peak during Stance????????????. .84 
Figure 25: Effect of Walking Velocity on rGM Peak during Stance???????????.. .85 
Figure 26: Effect of Walking Velocity on lGM Peak during Stance???????????.. .86 
Figure 27: Effect of Walking Velocity on ES Peak during Stance????? ????...??...87  
Figure 28: Effect of Cadence on Velocities????????????????????. ..90 
Figure 29: Effect of Cadence on Normalized Stride Length?????????????... ..91 
Figure 30: Effect of Walking Cadence on TP mean????????????????? ? .92 
Figure 31: Effect of Walking Cadence on TP SD?????????????????? ...93 
Figure 32: Effect of Walking Cadence on TL mean????????????????? ? 94 
Figure 33: Effect of Walking Cadence on PL mean????????????????? ? 95 
Figure 34: Effect of Cadence on Pelvic Rotation????????????? ???? ? .96 
Figure 35: Effect of Walking Cadence on Trunk Rotation?????????????... ...97 
Figure 36: Effect of Walking Cadence on Leg Rotation:?????????????? ? .98 
Figure 37: Effect of Cadence on PerL Peak???????????????????? ...100 
Figure 38: Effect of Walking Cadence on SOL Peak?????????????????.101  
Figure 39: Effect of Walking Cadence on BF Peak????????????????..?.102  
Figure 40: Effect of Cadence on rGM Peak????????????????????...10 3 
Figure 41: Effects of Cadence lGM Peak??????????????????? ??...10 4 
Figure 42: Effects of Cadence ES Peak????????????...?????????... 105 
Figure 43: Effect of Footwear on Normalized Stride Length during Normal Walking???...1 08 
Figure 44: Effect of Footwear on Velocity during Normal Walking??????????...109  
Figure 45: Effect of Footwear on Cadence during Normal Walking??????????...109  
xiv 
 
 
Figure 46: Effect of Gender on Walking Velocity during Normal Walking When Footwear is  
Altered? ??? ?????????????????????????????... ...110 
Figure 47: Effect of Gender on Cadence during Normal Walking When Footwear is Altered..111 
Figure 48: Effect of Footwear on PL mean???????????????????? ? 112 
Figure 49: Effect of Gender on TP SD?????????????????????... ...113 
1 
 
Chapter I. Introduction 
 For many individuals, walking is a fundamental means of moving from one location to 
another. Walking also serves as the foundation for other higher-level motor skills (e.g. running, 
skipping), providing a base for individuals to build more advanced skillsets (Haywood & 
Getchell, 2009). It is also a convenient and versatile means of locomotion, owing in large part to 
the functional versatility of the lower extremity and lumbopelvic regions. While walking is often 
thought of as a secondary motor task, a significant amount of resources have been spent on 
understanding normal and pathological gait (Perry, 1992; Whittle, 2003). Specifically, changes 
in gait economy and variability associated with neural deficits, altered postures, musculoskeletal 
limitations, and footwear are areas that continue to be examined by researchers (Perry, 1992; 
Smidt, 1990; Whittle, 2003). Furthermore, the aforementioned factors are capable of altering 
limb motion patterns associated with timing of musculature, thereby reducing economy and 
increasing the likelihood of cumulative trauma disorders (Bird, Bendrups, & Payne, 2003; 
Braunstein, Arampatzis, Eysel, & Bruggemann, 2010; Crosbie & Vachalathiti, 1997; Crosbie, 
Vachalathiti, & Smith, 1997a, 1997b; Divert, Mornieux, Baur, Mayer, & Belli, 2005; Divert et 
al., 2008; Gefen, Megido-Ravid, Itzchak, & Arcan, 2002; Y. Huang et al., 2010; Y. P. Huang et 
al., 2011; C. J. Lamoth, Beek, & Meijer, 2002; C. J. Lamoth, Meijer, et al., 2002; C. J. C. 
Lamoth, Daffertshofer, Meijer, & Beek, 2006; Maki, Perry, Norrie, & McIlroy, 1999; Murley, 
Landorf, Menz, & Bird, 2009; W. Wu et al., 2002; W. Wu et al., 2004; W. H. Wu et al., 2008). 
Footwear, in partially fulfilling its intended purpose, has been shown to significantly alter 
posture and gait mechanics when compared to barefoot gait; however, kinematic and kinetic 
changes associated with footwear are not always positive as the intended design would suggest 
(Lieberman et al., 2010; Robbins & Gouw, 1990; Sato, Fortenbaugh, & Hydock, 2012; Shakoor, 
2 
 
Lidtke, Sengupta, Fogg, & Block, 2008; Shorten & Mientjes, 2011). Recently, researchers have 
suggested that, despite significant improvements in footwear technology and materials science 
within the last several decades, kinematic and kinetic changes associated with shod gait may be a 
contributing factor to increased lower extremity injury rates (Lieberman et al., 2010; van Gent et 
al., 2007).  
When focusing on the lower extremity, researchers have found that 37% ? 79% of all 
individuals that regularly take part in weight-bearing physical activity develop a lower extremity 
injury, with a recurrence rate of 20% - 70% (van Gent et al., 2007). Likewise, the U.S. 
Department of Labor has repeatedly found that sprains, strains and tears are the dominant type of 
occupational injury for all types of workers, with injuries to the low back and lower-extremity 
accounting for almost 60% all workplace related injuries. In addition, while many individuals 
often associate a low back or lower-extremity injury with a specific event, epidemiological 
research has shown that most injuries result from a cumulative response over a considerable 
period of time (Andersson, 1998; Kraus, Schaffer, McArthur, & Peek-Asa, 1997; Magee, 2008; 
Marras, 2000; Marras et al., 1995; McGill, 2007). Therefore, it is not altogether surprising that 
four of the top five occupations with the highest injury rates also contain high volumes of 
physical activity and significant amounts of standing/ walking (laborers and manual material 
handlers; nurses or other healthcare individuals; janitors and cleaners, and lightweight delivery 
services).  
Another common concern within ergonomics is the role of awkward postures, or 
positions which significantly deviate from neutral or normal during a task, in the development of 
cumulative trauma disorders, a common characteristic for the previously mentioned occupations 
(M. Jackson, Solomonow, Zhou, Baratta, & Harris, 2001; Marras, 2000; Marras et al., 1995; Xu, 
3 
 
Bach, & Orhede, 1997). Though the ergonomics field does not typically view awkward postures 
as directly affecting the role of active and passive tissues during task completion, there is support 
that awkward limb positioning will reduce the capability of a tissue to tolerate work (Aultman, 
Drake, Callaghan, & McGill, 2004; Aultman, Scannell, & McGill, 2005; M. Jackson et al., 
2001). Therefore, it has been suggested that awkward postures will, by altering the behavior of 
the fatigue failure curve, reduce the number of cycles to failure (e.g. development of pain or 
injury) of a particular tissue. As a result, awkward postures have been implicated as a 
compounding factor in the development of cumulative trauma disorders.  
Research from the human performance field suggests that awkward postures may be 
directly involved in altering the response and loading patterns of active and passive tissues 
(Burton et al., 2005; Gallagher, Marras, Litsky, & Burr, 2005; N. D. Jackson, Gutierrez, & 
Kaminski, 2009; Kibler, 1998). The kinetic chain model of movement is based on the 
supposition that the proper sequencing of muscle activity patterns occurs in a proximal-to-distal 
nature, which thereby determines the level of movement economy and coordination. 
Furthermore, the kinetic chain model emphasizes the existence of a neuromuscular linkage, in 
which movements are initiated by torso musculature, followed closely by the extremities 
(McMullen & Uhl, 2000; Oliver, Dwelly, Sarantis, Helmer, & Bonacci, 2010; Oliver, Stone, & 
Plummer, 2010).  Therefore, the ideal firing and recruitment of the involved musculature is key 
to maintaining optimal biomechanical patterns during a movement and avoiding awkward 
postures. Within a sports and clinical setting, it has been postulated that detriments in muscular 
firing patterns, particularly those within the lumbopelivc hip complex (e.g. core), are 
contributing factors in a variety of injuries, such as shoulder complex pathology, low back pain 
and ACL injuries, among others (McGill, 2007; Willson, Dougherty, Ireland, & Davis, 2005). 
4 
 
Furthermore, other research has found that individuals with a history of lateral ankle sprains 
displayed altered firing patterns of the hip extensors on the ipsilateral and contralateral sides, 
indicating a significant relationship between lower extremity mechanoreceptors and torso 
musculature (Bullock-Saxton, Janda, & Bullock, 1994).  
While not yet completely understood, it is believed that the early onset of muscle 
activation within the core is necessary to improve the strength of the system, thereby providing a 
stable and reliable base from which more complex movement patterns can begin (Kibler, 1998). 
Similarly, others have found that the transverse abdominis, followed almost simultaneously by 
the multifidus, fire prior to any other limb movement or individual-initiated adjustments (Hodges 
& Richardson, 1997). In addition to the aforementioned muscles, more recent results have 
indicated that erector spinae activity also occurs prior to the onset of extremity excursion 
(Moseley, Hodges, & Gandevia, 2002). Therefore, it has been postulated that injuries or 
equipment which alter the mechanical and physiological variables involved with a movement 
will also considerably raise the associated injury risk. 
Despite the often perceived passiveness of footwear during movement, there is support 
within the literature to indicate that footwear contributes to awkward posture, resulting in 
significant physiological and mechanical alterations during movement. For instance, Sato and 
colleagues (2012), examining the effects of traditional running shoes and weightlifting footwear 
on barbell squat kinematics, reported significant differences in lower extremity and trunk 
kinematics. Specifically, participants displayed significantly greater trunk lean and trunk 
displacement when wearing traditional running shoes. Likewise, Shakoor et al. (2008) reported 
an 8-12% reduction in the peak external knee adduction moment, a contributing factor to the 
development of knee osteoarthritis, when participants wore specialized footwear as opposed to 
5 
 
self-chosen or control footwear. Aghazadeh and Liu (1994) reported a significant reduction in 
the lifting capacity of participants when tasks were performed in footwear with elevated instead 
of flat heels. Similarly, Opila-Correia (1990) reported significant changes in spatiotemporal 
variables and joint kinematics during high-heeled gait relative to low or flat-heeled walking. 
Specifically, it was reported that participants walked significantly slower, exhibited shorter stride 
lengths, increased knee flexion at foot strike and during stance phase, and displayed a reduced 
range of pelvic motion in the sagittal plane during high-heeled gait. More recently, it has been 
reported that high-heeled walking significantly reduced the ankle plantarflexor moment during 
stance, followed by a significantly increased hip flexor moment during the subsequent swing 
phase (Esenyel, Walsh, Walden, & Gitter, 2003). Clinically, footwear manipulations, through 
rehabilitative interventions or orthotic prescription, have routinely been utilized to significantly 
alter posture and movement patterns. (Gross, Davlin, & Evanski, 1991). 
There is also support within the motor behavior literature for understanding the role of 
footwear in gait mechanics. In 1986, Karl Newell suggested that human movement results from 
the interaction of the individual with the environment in which the movement occurs and the task 
to be undertaken (Haywood & Getchell, 2009). Furthermore, any variability in the three 
aforementioned factors will result in changes to the movement pattern. Therefore, the three 
constraints (individual, environmental, and task) create a dynamic model for understanding the 
development and execution of particular movement (Haywood & Getchell, 2009). The appeal of 
Newell?s Model of Constraints is that emphasis is placed on the influence of where the 
individual moves and what the individual does during particular movements. For example, it has 
been argued that while footwear choice may not considerably alter a carefully controlled walk, 
increasing the environmental or task demands would notably alter gait patterns (Haywood & 
6 
 
Getchell, 2009). As a result, a cumulative response of altered movement patterns will lead to 
changes in long-term motor behavior. Through its protection of the dorsal and plantar surfaces of 
the foot, footwear could be considered a necessary equipment choice for successful translation. 
Because the foot is the only extremity in direct contact with the ground during normal gait, 
proper functionality is vital to the execution of the chosen movement. Likewise, any equipment 
which alters the behavior of the lower extremity would lead to altered motor patterns. Therefore, 
footwear type would appear to be intricately linked to the dynamics of any movement including 
walking.  
Understanding the kinematic differences in barefoot versus shod gait may allow for a 
greater understanding of the kinetic differences. Saunders and colleagues (1953) were one of the 
first to attempt to classify the movements of limbs during gait, resulting in the formation of a 
system collectively referred to as the Six Determinants of Gait. Of particular interest to the 
proposed research is the pelvic rotation, which primarily occurs during late swing and terminal 
stance. It is believed that the pelvic rotation allows for a functional leg length increase during 
ipsilateral swing, reducing hip amplitude range of motion and center of mass excursion, thereby 
decreasing the physiological costs of walking. However, unlike the linear relationship between 
running velocity and oxygen consumption, it has been found that there is a curvilinear 
relationship between these variables during walking (Margaria, 1976). The increased 
physiological cost associated with walking at higher velocities is likely related to an increased 
stride length and related changes in mechanical variables (e.g. anterior-posterior forces) (Divert 
et al., 2005; Divert et al., 2008; Margaria, 1976). While walking velocity is the product of stride 
length and stride rate (e.g. cadence), increasing stride rate appears to be the predominant factor in 
increasing walking velocity during barefoot translation (Divert et al., 2008; Lieberman et al., 
7 
 
2010; Romer et al., 2012). Conversely, recent research suggests that increasing stride length is 
the predominant factor to increasing walking velocity during shod gait (Divert et al., 2008; 
Lieberman et al., 2010).  
The alteration of the aforementioned spatiotemporal variables may have considerable 
implications for the pelvic step. As described by Ducroquet et al. (1968), the pelvic step is a 
transverse pelvic girdle rotation, acting to lengthen the step as velocity increases. It is generally 
accepted that there is a positive relationship between walking velocity and the magnitude of the 
pelvic step for velocities greater than 0.8 ms-1  (Bruijn, Meijer, van Dieen, Kingma, & Lamoth, 
2008; Crosbie & Vachalathiti, 1997; Crosbie et al., 1997a; Y. Huang et al., 2010; C. J. Lamoth, 
Beek, et al., 2002).  Furthermore, Ducroquet et al. (1968) asserted that the counter-rotation of the 
thorax is initiated to limit total angular momentum about the vertical axis during gait; however, 
more recent findings question the thorax and pelvic contributions to total body angular 
momentum during walking (Bruijn et al., 2008). Therefore, it has been argued that the current 
knowledge regarding the understanding of the pelvic step is poor, necessitating an updated 
approach to understand the mechanisms responsible for pelvic step alterations (Bruijn et al., 
2008). For instance, yet to be understood is the effect of walking velocity, stride length, and 
cadence on the relative timing and magnitude of thorax, pelvic, and hip movements during gait.  
 As previously alluded to, a near universal finding of footwear research is an increased 
step length during shod locomotion, regardless of walking velocity or cadence (Nurse, Hulliger, 
Wakeling, Nigg, & Stefanyshyn, 2005; Romer et al., 2012; Wegener, Hunt, Vanwanseele, Burns, 
& Smith, 2011). It had been hypothesized that observed changes in shod gait kinematics are the 
result of mechanical factors (i.e. increased inertia of the distal segment); however, other authors 
have hypothesized a neural cause (e.g. reduced afferent feedback) as the root of alterations in 
8 
 
shod gait mechanics  (Divert et al., 2008; Kurz & Stergiou, 2003; Mundermann, Stefanyshyn, & 
Nigg, 2001; Murley et al., 2009; Nurse et al., 2005). Specifically, Kurz and Stergiou (2003) 
found that gait variability can be significantly affected by peripheral sensory information. 
Furthermore, it has been suggested that tactile feedback is an important control mechanism for 
maintaining balance and locomotion, with authors finding augmented plantar facilitation capable 
of improving balance during static trials and altering muscle activity during gait (Bird et al., 
2003; Maki et al., 1999; Nashner, 1980; Nurse et al., 2005). Likewise, Kurz and Stergiou (2003) 
found the reduced sensory information associated with shod gait resulted in a significant 
reduction in kinematic variability as compared to barefoot locomotion. Therefore, it appears that 
footwear mutes or reduces the tactile and proprioceptive feedback received by the lower 
extremity, resulting in altered gait mechanics.  
Support for a neural interaction in gait mechanics has also arisen from areas of research 
such as peripheral neuropathy, textured insoles, and balance. For instance, there is evidence that 
footwear reduces the tactile and proprioceptive feedback of the lower extremity, resulting in an 
altered response  (i.e. timing and magnitude) of lower extremity and torso musculature during 
standing and gait (Allet et al., 2009; Bird et al., 2003; Maki et al., 1999; Murley et al., 2009; 
Nawoczenski & Ludewig, 1999; Nurse et al., 2005; Nurse & Nigg, 2001; Wakeling & Liphardt, 
2006; Wakeling, Liphardt, & Nigg, 2003). Nurse et al. (2005) found significant differences in the 
magnitude of lower leg muscle activity (i.e. soleus, tibialis anterior) during walking when using a 
textured versus smooth insert. Furthermore, Bird et al. (2003) reported significant differences in 
the timing of torso and hip musculature (i.e. erector spinae, gluteus medius) during walking with 
varying foot wedges. Relative to barefoot gait, shod gait most often displays an increased stride 
length, increased anteroposterior ground reaction forces (GRF), and increased vertical GRF 
9 
 
impact transient, changes which may assist in increasing sensory feedback of the foot during 
locomotion (Kurz & Stergiou, 2003; Lieberman et al., 2010). 
Beyond footwear research, support for the neural interaction in gait mechanics can be 
found in the mechanical and physiological changes during the gait of pathological populations 
(e.g. peripheral neuropathy) or healthy individuals with an induced reduction in plantar sensation 
(Manor & Li, 2009; Savelberg et al., 2010). Consequently, these findings suggest that footwear 
may result in considerable differences in the loading of the passive and active tissues during gait. 
If it is possible to cause a meaningful change in the efferent response of the active musculature 
through an altered afferent response, greater loading of the associated elastic tissues may spare 
the passive tissues from considerable microtrauma.  
Summary 
 In conclusion, there is considerable evidence that footwear significantly alters gait 
mechanics (Braunstein et al., 2010; Kristen, Kastner, Holzreiter, Wagner, & Engel, 1998; 
Lieberman et al., 2010; Wegener et al., 2011). Specifically, authors have found an increased 
stride length, decreased cadence, increased impact transient, and altered joint kinematics (e.g. 
increased hip flexion, decreased plantarflexion) during shod locomotion (Braunstein et al., 2010; 
Kristen et al., 1998; Lieberman et al., 2010; Wegener et al., 2011). Physiological changes have 
also been found during shod gait, including a reduced metabolic economy and altered or delayed 
muscular activation of lower leg and torso musculature (Bird et al., 2003; Divert et al., 2008; 
Maki et al., 1999; Murley et al., 2009; Nurse et al., 2005; Nurse & Nigg, 2001). Furthermore, 
footwear has been implicated in altered lumbar and pelvic positioning while standing and may 
even alter the response of lower extremity and lumbar musculature (Aghazadeh & Lu, 1994; 
Chiu & Wang, 2007; Cikajlo & Matjacic, 2007; Keenan, Franz, Dicharry, Della Croce, & 
10 
 
Kerrigan, 2011; C.-M. Lee, Jeong, & Freivalds, 2001). The combined findings of the Kinetic 
Chain model and Newell?s Model of Constraints provide a basis for the role of footwear leading 
to altered movement patterns during gait. Specifically, the role of footwear on mechanical and 
physiological variables during gait must be further delineated to provide an understanding of 
injury pathways and future product innovation. The relationship between trunk, pelvic, and lower 
extremity movements during gait is important in maintaining forward momentum and increasing 
economy; however, the relevance of the timing of the segmental movements is not yet fully 
understood. While countless studies have examined the role of footwear on walking and running 
mechanics, only one study has examined the effect of spatiotemporal variables on lumbopelvic 
rhythm during gait (Y. Huang et al., 2010). However, to the knowledge of the author, no study 
has simultaneously examined the role of footwear on limb segment timing and muscle activity of 
the lumbopelvic and lower-extremity areas, nor the effect of independently manipulating 
spatiotemporal variables (i.e. velocity, cadence) on thorax-pelvic coordination. To advance the 
understanding of human movement, the present project examined the effect of spatiotemporal 
variables and footwear manipulation on trunk coordination and muscle activation during 
walking.  
Purpose of the Study 
 The purposes of this investigation was multi-faceted: (1) to investigate the effects of 
velocity and cadence on lumbopelvic rhythm during overground walking; (2) to investigate the 
effects of footwear on lumbopelvic rhythm during gait; (3) to examine the effects of enhanced 
tactile feedback (i.e. barefoot, textured insole) on the onset of muscle activity at foot strike; (4) to 
examine the effects of spatiotemporal and footwear variation on the magnitude of lumbar (i.e. 
erector spinae) and lower extremity muscular activity (e.g. gluteus medius, biceps femoris, 
11 
 
soleus, peroneus longus) during the stance phase of gait, and (5) to investigate differences in the 
kinematics and electromyography between males and females.  
Hypotheses 
The null hypotheses for the present study are as listed: 
H01: Velocity will not have a significant impact on the magnitude or timing of 
lumbopelvic rhythm during overground walking. 
H02: Cadence will not have a significant impact on the magnitude or timing of 
lumbopelvic rhythm during overground walking. 
H03: There will not be any significant differences in lumbopelvic rhythm between 
barefoot and shod overground walking. 
H04: Enhanced tactile feedback will not result in a significantly different onset of 
muscular activation for either lumbar or lower extremity musculature.  
H05: Males and females will not differ in kinematics or electromyography during 
overground walking.  
Limitations 
Limitations for the present study are: 
1. Self-reported health status will be used. 
2. Not every type of footwear will be evaluated. 
3. Participants will not be demarcated based on the amount of barefoot gait experience. 
4. The age of participants will be restricted to individuals younger than 35 years of age. 
Delimitations 
The delimitations for the present study are: 
1. Participants will be required to wear retroreflective markers bilaterally.  
12 
 
2. Surface electrodes for the peroneus longs, soleus, and biceps femoris will be collected 
unilaterally (right side of body) while the gluteus medius will be collected bilaterally.  
3. Individuals with chronic low-back pain will be excluded from participation.  
4. The footwear and inserts that will be utilized for this study will be new.  
5. The participants will be assessed in the Auburn University Lower Extremity 
Laboratory which contains space specifically designed for gait-related research.  
Definition of Terms 
Active Tissues 
 The musculoskeletal tissues that directly contribute to human movement and participate 
in internal force production (e.g. muscles). 
Kinematics 
 Kinematics is a branch of mechanics that is involved in the description of movement 
without regard to the forces that have caused the movement. Variables include linear and angular 
displacement, velocity, and acceleration. 
Kinetics 
 Kinetics is a branch of mechanics that focuses on the internal and external forces that 
have caused a movement. 
Kinetic Chain 
 The kinetic chain model is based on a network of individual links that, through mutual 
dependency and sequencing of muscle activity patterns in a proximal-to-distal nature, allow for 
the execution of coordinated and efficient movements. 
 
 
13 
 
Lumbopelvic Rhythm 
 Refers to the postural relationship between the pelvis and lumbar spine during static and 
dynamic movements. 
Passive Tissues 
 The tissues of the body which may passively produce force, restrict motion, modify the 
direction of the applied force, or provide support, yet do not take part in active force production. 
Traditionally, these tissues would be regarded as fascia, ligaments, and various retinacula.  
Relative Phase 
 The timing relationship between segments during walking is referred to as the relative 
phase, which provides a measure of the coordination between two segments during a movement.  
Surface Electromyography 
Electromyography is the study of electromyograms (EMG), the electrical signal 
associated with muscular contraction. Surface electromyography (sEMG) is a technique for 
evaluating and recording the voluntary muscle activity. Finally, an electromyograph is 
representative of the spatiotemporal summation of motor unit action potentials (MUAP) within 
the capture volume of the electrode.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
14 
 
Chapter II. Review of Literature 
 Walking has been the fundamental means of locomotion for humans for thousands of 
years, providing a convenient means of short-term transportation. Similarly, various forms of 
footwear have been used throughout mankind?s history, however, the burgeoning footwear and 
fitness market of the 1970?s resulted in a significant transformation in the design and materials 
utilized in modern footwear (Bramble & Lieberman, 2004; Lieberman et al., 2010; McDougall, 
2009). This study will build upon previous literature examining the influence of footwear on gait 
mechanics and the role of lower extremity, pelvic, and thorax movements to better understand 
movement pattern changes that may lead to low back or lower-extremity pathology. The 
purposes of this investigation was multi-faceted: (1) to investigate the effects of velocity and 
cadence on lumbopelvic rhythm during overground walking; (2) to investigate the effects of 
footwear on lumbopelvic rhythm during gait; (3) to examine the effects of enhanced tactile 
feedback (i.e. barefoot, textured insole) on the onset of muscle activity at foot strike; (4) to 
examine the effects of spatiotemporal and footwear variation on the magnitude of lumbar (i.e. 
erector spinae) and lower extremity muscular activity (e.g. gluteus medius, biceps femoris, 
soleus, peroneus longus) during the stance phase of gait, and (5) to investigate differences in the 
kinematics and electromyography between males and females.  
This chapter is divided into five sections. Section one describes human gait and the 
various phases of walking. Section two describes the role of lumbopelvic region in human 
locomotion. Section three describes the influence of footwear on gait mechanics. Section four 
describes the role of plantar sensation and tactile feedback in human locomotion. Section five 
summarizes the pertinent findings of the previously discussed literature, emphasizing its 
relevance to the present study.  
15 
 
Section 1: Human Gait and the Phases of Walking 
 To differentiate between walking and other gait variations (i.e. running), walking can be 
defined as involving the use of two alternating legs, providing both support and propulsion, in 
which one foot is in contact with the ground at all times, and often accompanied by a period of 
double support (Perry, 1992; Smidt, 1990; Whittle, 2003). While gait can refer to a variety of 
movements (e.g. walking, running, trotting), for the purposes of this paper, gait will be in 
reference to the normal, bipedal pattern of forward locomotion. A complete gait cycle is one in 
which a limb cycles from a period of support, to non-support, to ipsilateral support. Because of 
the cyclic nature of gait, this cycle is completed alternately by each of the lower limbs in 
continuous fashion. Perry (1992) has outlined three basic approaches to gait analysis: (1) division 
of the gait cycle based on the reciprocal contact of the two lower extremities; (2) the utilization 
of spatiotemporal variables to describe the gait cycle; and (3) the examination of the various 
components of the gait cycle based on the functional significance of each event. This section 
provides an overview of each of the aforementioned gait analysis approaches. Furthermore, 
information concerning the phases and functions of the gait cycle are also described.  
Division of Gait Cycle based on the Reciprocal Contact of the Lower Extremity 
 The simplest approach to gait analysis is a division based upon the reciprocal patterns of 
ground contact by the feet (Perry, 1992; Whittle, 2003). During walking, at any point in time, 
one limb serves as the base of support while the other limb is in motion to the next support site. 
As previously mentioned, a complete gait cycle involves a period of support and non-support for 
the same limb. Because of the cyclic nature of walking, this process continually repeats itself, 
with no true ending or starting point, until the destination is reached. In traditional gait analysis, 
the moment of initial foot contact has been utilized as the start of the gait cycle. The term initial 
16 
 
contact is utilized as the generic term due to variations in foot strike patterns between 
pathological or types of gait (e.g. running). Within this approach, the gait cycle can be divided 
into two distinct periods: stance phase and swing phase. The stance phase is used to designate the 
entire time period that the limb is in ground contact, beginning at initial contact and ending at 
toe-off. The swing phase encompasses the period of time in which the limb is airborne, 
beginning at toe-off and ending at initial contact, for the purposes of limb advancement (Figure 
1). 
 
Figure 1. Phases of Gait (Kirtley, 2008). 
 
Division of Gait Cycle based on Spatiotemporal Variables 
 The second approach to analyzing gait builds upon the foundation of the first approach, 
but further divides the sequencing of gait events based on the temporal relationship between the 
lower limbs (Perry, 1992). Specifically, the stance phase is divided into initial double stance 
phase, single limb support, and terminal double stance phase. Initial double stance occurs after 
initial contact (ipsilateral initial contact), signifying the beginning of the gait cycle. Though there 
17 
 
is some load sharing between limbs during this phase, the forward limb is in the initial stages of 
weight acceptance while the rear leg is beginning limb advancement. Single limb support occurs 
when the contralateral foot ceases ground contact (contralateral toe-off) for the swing phase. As 
the name suggests, there is no load sharing between limbs and the entire weight of the body is 
supported by the stance leg. Terminal double support stance is most begins at the instant of 
ground contact by the swing leg (contralateral initial contact), continuing until the original stance 
limb ceases ground contact (ipsilateral toe-off). Because of the cyclic and offset nature of 
walking, the initial double stance phase for one limb corresponds to contralateral terminal double 
stance (Figure 2).   
  
Figure 2. Functional Phases of Gait (Kirtley, 2008). 
  
Normal walking is typically associated with a 60% stance phase and 40% swing phase for 
each limb, though this relationship is dependent on velocity. Specifically, as walking velocity 
increases, there is an increased proportionality of the swing phase, with the stance phase 
becoming shorter. Perry (1992) has reported that the initial double stance period accounts for up 
to 10% of the gait, or about one-sixth of the overall stance phase (Table 1). Approximately 40% 
of the gait cycle, or two-thirds of the stance phase, is spent in single-limb support. Another 10% 
18 
 
of the gait cycle, or one-sixth of the stance phase, is spent in terminal double stance as the stance 
leg prepares for limb advancement (40% of the gait cycle).  
Table 1 
Percentages of stance and swing phases during gait (Perry, 1992). 
 Ground Contact 
Stance 60%  
- Initial Double Stance  10% 
- Single Limb Support  40% 
- Terminal Double Stance  10% 
Swing 40%  
 
 
 As previously cited, the 60% / 40% is a rough estimate and is based upon the normal 
walking velocity (a function of stride length and cadence). Blanc and colleagues (1999) 
demonstrated that healthy adults exhibited a stance phase of ~62% and a swing phase of ~38% 
when walking barefoot at a preferred velocity (values not reported). However, it has previously 
been reported that there is a curvilinear relationship between phase times and walking velocity 
(Murray, Drought, & Kory, 1964). Specifically, as walking velocity increases, there is a decrease 
in the initial and terminal double stance phase intervals, yet a concurrent increase in single-limb 
support. Therefore, as walking velocity increases, an individual begins to spend a greater 
percentage of time in single limb support, until ultimately the double stance phase is entirely 
eliminated. The disappearance of a period of double support signifies the transition from walking 
to running as there is now a period of non-support or a flight phase (Blanc, Balmer, Landis, & 
Vingerhoets, 1999; Perry, 1992; Whittle, 2003). Finally, as walking velocity decreases, there is 
an increase in the duration of the stance phase. Conversely, as walking velocity increases there is 
an increase in the duration of the swing phase.  
 Additional descriptive terms exist to describe the placement of the lower limbs within a 
gait cycle (Perry, 1992). The stride length is the distance between two successive placements of 
19 
 
the ipsilateral foot and is often supplanted by step length; however, step length refers to the 
distance between the initial contact of one foot to the initial contact of the contralateral foot. 
Therefore, stride length is composed of two successive steps. In the absence of a pathological 
condition, step length has been found to be fairly symmetrical, with each step contributing 
approximately the same distance to the overall stride length (Perry, 1992; Whittle, 2003). The 
walking base, also known as step width or base-of-support, is the side-to-side distance between 
two steps, typically measured from the mid-posterior points of the heel. During normal gait there 
is an inverse relationship between walking velocity and step width, with increasing velocity 
associated with a decreasing step width. 
Division of Gait Cycle using Functional Significance of Events 
 While the aforementioned divisional components provide a basis for understanding the 
gait cycle, researchers have incorporated additional divisions which more readily identify the 
functional demands of each phase of gait (Perry, 1992; Smidt, 1990; Whittle, 2003). This system 
has resulted in the identification of eight distinct sub-phases for each stride (Figures 2 & 3). 
Retracing the previous steps, the gait cycle can be divided into stance and swing phases. It has 
been further divided into periods of initial double stance, single limb stance, terminal double 
stance, and swing phase. However, these limited descriptions do little to explain the functional 
significance and contribution of the pelvic and lower extremity joints to the overall movement. 
Hence, the utilization of analyzing walking patterns through sub-phases aids in understanding the 
contribution of a particular phase to the overall movement (Perry, 1992). Furthermore, this 
approach aids in understanding pathology during gait. A posture or movement appropriate for 
one phase would create dysfunction in another phase. Therefore, based on the task demands of 
20 
 
gait cycle, the stance phase is sub-divided into weight acceptance and single limb support phases. 
Likewise, the purpose of the swing phase is for limb advancement.  
 
 
 
Figure 3. Gait Cycle Divisions (Perry, 1992). 
 
The onset of the stance phase signifies weight acceptance, which includes two additional 
phases: initial contact and loading response. Single limb support is the next component which is 
made up of the mid-stance, terminal stance, and pre-swing phases. The final component of the 
gait cycle is limb advancement, made-up of pre-swing, initial swing, mid-swing, and terminal 
swing phases. Pre-swing has been determined to be a component of the single limb and limb 
advancement phases due to the importance of this component and because postural changes 
begin during the stance phase to meet the high demands of limb advancement (Figure 3).  
Weight acceptance is often considered the most demanding task of the gait cycle, 
coordinating movements for shock absorption, initiating limb stability, and maintenance of 
forward progression. Furthermore, the weight acceptance task is involved with the initial contact 
and loading responses phases. Initial contact, accounting for 0-2% of the gait cycle, begins the 
instant of foot contact. Joint positioning during this phase is a determining factor of the limb?s 
loading response (Perry, 1992). Another purpose of this phase is to position the lower limb and 
provide a starting point for the stance phase. This allows for the maintenance of the forward 
21 
 
momentum of the body, often referred to as the heel rocker. The second phase is the loading 
response, accounting for 0-10% of the gait cycle (Perry, 1992; Whittle, 2003). This phase 
corresponds to the initial double stance period, beginning with initial contact and concluding at 
contralateral toe-off. The loading phase is multi-functional serving the purposes of shock 
absorption, establishing stability of the limb for weight-bearing, and preserving the forward 
progression of the body.  
The task of single limb support involves the functional phases of mid-stance, terminal 
stance, and pre-swing. The purpose of the single limb support task is to support the body with the 
stance leg while the contralateral limb advances to the next site of foot contact. Mid-stance 
accounts for approximately 20% of the overall gait cycle, or the time period from 10 ? 30% of 
the total gait cycle interval (Perry, 1992; Whittle, 2003). This phase begins at contralateral toe-
off (e.g. contralateral swing), continuing until the body is situated over the forefoot. While the 
mid-stance phase assists in maintaining the forward momentum of the body, it also aids in limb 
and trunk stability. Terminal stance, accounting for another 20% of the overall gait cycle, occurs 
over the 30-50% interval of the total gait cycle (Perry, 1992; Whittle, 2003). While the primary 
objective of the terminal stance phase is similar to that of mid-stance (maintain forward 
progression of the body), the difference is the manner in which the forward progression is 
maintained. Specifically, during terminal stance, which begins at heel-off and continues until 
foot strike of the contralateral limb, the center of mass of the body is anterior to the forefoot. 
Therefore, rotation occurs about the forefoot rocker, aiding in the considerable force production 
of the plantarflexors (Whittle, 2003).  
Limb advancement is the final task of the gait cycle, constituting approximately 50% of 
the overall gait cycle (Perry, 1992). While pre-swing is formally included in this section, it is in 
22 
 
fact a carryover from postural adjustments that begin during stance and is synonymous with the 
terminal double stance interval. During normal gait, the pre-swing phase is approximately 10% 
of the overall gait cycle, covering the 50-60% interval of the gait cycle (Perry, 1992; Whittle, 
2003). Like terminal double stance, this phase begins at initial contact of the contralateral limb, 
ending at ipsilateral toe-off. The primary purpose of this phase is to unload the limb so the 
postural and muscular actions are more readily able to contribute to latter phases. The sixth phase 
of the gait cycle is initial swing, accounting for one-third of the overall swing period (60-73% of 
the gait cycle interval) (Perry, 1992; Whittle, 2003). Beginning at ipsilateral toe-off, the initial 
swing phase ends when the swing foot is in-line with the contralateral stance foot. This phase is 
synonymous with mid-stance of the contralateral limb. Initial swing serves to aid in clearing the 
swing foot of the floor or other ground obstacles while advancing the limb from the trailing 
position. Mid-swing is the seventh phase of the gait cycle, constituting the interval of 73-87% of 
the total gait cycle (Perry, 1992; Whittle, 2003). Serving the same purposes of the initial swing 
phase, the mid-swing phase begins when the swinging limb is opposite the stance limb, ending 
when the lower leg of the swinging segment is perpendicular to the ground. The final phase of 
the gait cycle, terminal stance, accounts for the remaining 13% of the gait cycle (87-100% of the 
gait cycle interval) (Perry, 1992; Whittle, 2003). Beginning when the lower leg is vertical to the 
floor, it terminates upon foot contact of the swinging limb. Therefore, while limb advancement is 
served during this phase, it must also prepare the limb for weight acceptance at ground contact.  
Locomotor and Passenger Units 
 While much of the previous discussion has focused on the lower body, it could be 
considered remiss to so simply isolate the task of walking. To address the relative importance of 
various segments, researchers have divided the body into two functional units that represent 
23 
 
distinct movement patterns: the passenger unit and the locomotor unit (Perry, 1992). The head, 
neck, trunk, and arms comprise the passenger unit, often referred to as HAT. The locomotor unit 
consists of the lower limbs and pelvic area, including eleven articulations of interest during 
walking: lumbosacral, bilateral hip, knee, subtalar, and metatarsophalangeal joints. Comprising 
almost 70% of the total mass of the body, the HAT segments have been collectively referred to 
as the passenger unit due to the relative passiveness of the segments to the contribution of 
walking. Specifically, there is a low need for muscular activity beyond maintaining the segments 
in the chosen posture; however, there may still be considerable movement. For instance, arms 
swing out of phase with the legs. Therefore, the left leg, left pelvis, and right arm tend to rotate 
anteriorly while the right leg, right pelvis, and left arm rotate posteriorly. Despite the 
considerable movement, there appears to be minimal active contribution of the arms to overall 
gait economy during normal velocity. The primary purpose of the passenger unit is to maintain 
the balance and posture of the HAT, and center of mass of the body, which aid the lower limbs in 
maintaining the forward progression and base of support translations.  
Six Determinants of Gait 
 The locomotor unit is responsible for generating propulsive force, maintaining the upright 
stability of the body, aiding in shock absorption during ground contact, and conserving energy 
during translation. Controlling the large number of joints within the locomotor unit, 57 muscles 
must act in a controlled and coordinated manner to ensure successful execution of the multi-
segmental limbs. To aid in understanding the relationship between the passenger unit, locomotor 
unit, and walking economy, Saunders and colleagues (1953) identified the mixture of six motion 
patterns that are now known as the Six Determinants of Gait. Three of the determinants are 
related to the pelvic motions (pelvic rotation, contralateral drop, and lateral shift) while three are 
24 
 
focused on lower extremity movements (knee flexion during stance, ankle mechanism, and 
terminal rocker). While it was originally proposed that the primary purpose of the six 
determinants was to minimize center of mass excursion, and therefore energy expenditure during 
gait, this hypothesis has been debated in recent years (Croce, Riley, Lelas, & Kerrigan, 2001; 
Gard & Childress, 1997; Whittle, 2003). Nonetheless, the movements have been found to occur 
and are of clinical and research interest.  
Pelvic rotation is the first determinant of gait and the one of greatest relevance to the 
present project. Occurring as the swing limb advances through the swing phase, pelvic rotation 
twists about a vertical axis, thereby placing the hip joint of the swing leg anterior to that of the 
stance leg (Figure 4) (Saunders, Inman, & Eberhart, 1953; Whittle, 2003). Furthermore, the 
rotation can also be seen during terminal stance; however, the rotation during terminal stance 
occurs in the posterior direction. Pelvic rotation aids in accomplishing several tasks. First, pelvic 
rotation results in a functional increase in limb length as stride length increases. Second, the 
pelvic rotation assists in moving the hip joints closer to the midline of the body. Finally, pelvic 
rotation reduces the flexion-extension range of motion, relative to the distal extremity, necessary 
to achieve a given stride length. Another determinant is the contralateral drop that occurs during 
the swing phase of gait. Contralateral drop appears similar to a positive Trendelenburg test, but is 
not pathological in origin, and occurs about an anteroposterior axis when the swing limb lowers 
approximately 4? during the swing phase (Figure 5). The purpose of the contralateral drop is to 
aid in reducing the vertical translation of the HAT during stance phase.  
 
25 
 
 
Figure 4. Six Determinants of Gait: Pelvic Rotation (Ayyappa, 1997). 
 
 
Figure 5. Six Determinants of Gait : Contralateral Drop (Ayyappa, 1997). 
 
 
Lateral displacement of the pelvis is the third determinant of gait and consists of two 
factors (Figure 6) (Saunders et al., 1953; Whittle, 2003). First, the natural valgus arrangement 
between the femur and tibia aids in placing the feet closer to the middle of the body during 
ground contact. Second, the genu abduction that occurs during loading shifts the center of center 
26 
 
of mass towards the supporting foot. Another determinant is the terminal or forefoot rocker, 
which corresponds to heel rise of the stance limb (Figure 7). The purpose of heel rise is to aid in 
increasing the functional length of the leg, allowing the foot to shift the center of pressure 
towards the forefoot. The fifth determinant, the ankle mechanisms, acts similar to heel rise. 
However, instead of increasing the effective length of the limb during terminal stance, the ankle 
mechanism increase acts to lengthen the limb prior to initial contact. This effect is achieved 
through dorsiflexion at the talocrural joint, aiding initial contact by extending the foot during 
walking. The final and sixth determinant of gait is knee flexion during the stance phase (Figure 
8). As the hip passes from flexion to extension, concurrent knee flexion acts to reduce the 
vertical height between the center-of-mass and ground, thereby reducing vertical displacement.  
 
 
 Figure 6. Six Determinants of Gait: Lateral Displacement (Ayyappa, 1997). 
 
 
27 
 
 Figure 7. Six Determinants of Gait: Foot & Ankle Motion (Ayyappa, 1997). 
 
 
 
 Figure 8. Six Determinants of Gait: Knee Motion (Ayyappa, 1997). 
 
 
The previous discussion provides a foundation for a clinical and scientific approach to 
gait analysis. The present project will be concerned with understanding the effects of footwear 
on relationships between limb advancement, the passenger unit, and the locomotor unit. Of 
particular interest is the relationship between these variables and lumbopelvic movements during 
gait. Specifically, this project focused on: (1) stride length from initial contact of the right limb 
until subsequent initial contact of the ipsilateral limb; (2) the relative timing of leg, pelvic, and 
thorax movements; (3) the absolute amplitude of leg, pelvic, and thorax movements; and (4) 
sEMG activity of the erector spinae, gluteus medius, biceps femoris, soleus, and peroneus longus 
in conjunction with joint kinematics of the lumbopelvic region and lower extremity.  
Section 2: Role of the Lumbopelvic Region on Human Locomotion 
 Walking is the most common form of human locomotion and one of the most researched 
areas of functional human activity. The predominant focus of most gait research is on the lower 
extremity, with a secondary focus on the lumbopelvic complex. Historically, much of the 
28 
 
lumbopelvic research has relied on descriptive literature of spatiotemporal parameters to explain 
movements during gait. As discussed in Section 1, Saunders and colleagues (1953) developed a 
list of gait characteristics, now referred to as the Six Determinants of Gait. Three of the 
determinants describe specific movements by the pelvis, which aid in the maintenance of 
forward momentum and energy conservation. Likewise, Ducroquet et al. (1968) introduced the 
pelvic step concept, stating that from a certain walking velocity and faster, pelvic rotations are 
dominantly responsible for increases in step length and walking velocity (Ducroquet, Ducroquet, 
& Ducroquet, 1968). Though Ducroquet and colleagues described the relationship in very 
general terms, more recent research has suggested that the pelvic step becomes an important 
contributor to step length at velocities greater than 0.8 ms-1 (Bruijn et al., 2008; Wagenaar & 
Beek, 1992). 
Recently, the timing relationship (e.g. relative phase) between thorax and pelvis rotations 
during walking has received some attention within the literature (Bruijn et al., 2008; Y. Huang et 
al., 2010; C. J. Lamoth, Beek, et al., 2002; van Emmerik & Wagenaar, 1996). At all walking 
velocities, the thorax counter-rotates with respect to the upper legs; however the pelvic timing 
appears to change from in-phase (synchronous) to out-of-phase (asynchronous) as velocity 
changes. Remaining unanswered though is the mechanism underlying thorax-pelvis coordination 
during normal and pathological gait. Similarly, there are conflicting hypotheses as to the role of 
the vertebral column during gait. Some authors assert spinal motion occurs in response to the 
action of the lower limbs while other authors have argued for a ?spinal engine? generation of gait 
patterns (C. J. Lamoth, Beek, et al., 2002; Taylor, Goldie, & Evans, 1999; van Emmerik & 
Wagenaar, 1996). Finally, the author is unaware of any studies examining the effect of footwear 
on the lumbopelvic complex, with only minimal literature available regarding the footwear effect 
29 
 
on activity of trunk musculature during gait. Literature examining the relative phase (RP) 
relationship of the lumbopelvic complex is reviewed in greater detailed below. 
Several authors have examined relative phase differences during gait in healthy 
populations, with much of the research originating from only a handful of research laboratories. 
In one of the earlier papers of interest, van Emmerik and Wagenaar (1996) published a paper 
aimed at understanding the effects of walking on trunk motion. Seven young, male participants 
walked on a treadmill while gait kinematics were collected utilizing a motion capture (LED) set, 
operating at 60 Hz, and attached at the T6 level, pelvis, and a single upper leg. During van 
Emmerik and Wagennar?s (1996) study, participants walked at a pre-determined velocity, 
beginning at 0.3 ms-1 and ascending to 1.3 ms-1 in 0.1 ms-1 intervals before decreasing back to 
the original velocity. Furthermore, participants walked barefoot for the duration of the protocol. 
Not surprisingly, the authors found that stride length exhibited a significant increase as walking 
velocity increased. Interestingly though, the descending portion of the protocol exhibited a 
hysteresis effect, with a considerably larger stride length than during the ascending portion. 
Similar effects were found for stride frequency, however, a significant hysteresis effect was 
found with lower stride frequencies during the descending portion (van Emmerik & Wagenaar, 
1996). Likewise, while total range of motion (ROM) of pelvic rotation was greater during higher 
walking velocities, there was also greater pelvic rotation ROM during the descending portion of 
the protocol as compared to the ascending. Similar significant effects were found for thorax 
rotation. The authors reported that thorax-pelvis RP increased from approximately 25? at 0.3 ms-
1 to 110? at 1.3 ms-1. A similar decrease was witnessed during the descending portion of the 
protocol, however, no significant hysteresis effect was found. Together, the results support the 
hypothesis that thorax-pelvis rotations transition from more in-phase to more out-of-phase as 
30 
 
walking velocity increases. Remaining unanswered though is the mechanism responsible for 
these changes (e.g. stride length, cadence).  
A series of papers were published in 1997 by Crosbie et al. that examined age and gender 
effects on spinal, pelvic, and hip kinematics during gait. Unlike the aforementioned studies, these 
studies utilized a rather large sample (n=108), composed equally of each gender, and from a 
considerable age range (20 to 82 years of age). Also, participants completed all trials barefoot. 
The authors relied on the use of more traditional methodology (e.g. passive motion capture 
system) which included thirteen retroreflective markers on the dorsal surface of the thorax and 
lower extremity. Peak lateral trunk flexion was found to occur at 15% and 65% of the total gait 
cycle, corresponding to early swing phase. During this period, thorax and lumbar regions 
exhibited a displacement towards the stance leg while the pelvis shifted towards the swing leg. 
The authors also reported that the pelvis and trunk returned to a neutral posture during late swing 
or early double support. The authors noted a significant posterior pelvic girdle tilt at initial 
contact, followed by an anterior pelvic girdle rotation during the initial period of ground contact, 
before displaying a progressive posterior pelvic girdle tilt before the next foot strike. Likewise, 
maximum lumbar flexion was found to occur at heel strike, with a return to neutral during single 
limb support, before exhibiting flexion prior to contralateral foot strike. Finally, the authors 
found that lumbar spine motion tended to follow pelvic movements, though exhibited a slight 
phase lag.  
 The second paper by Crosbie and colleagues (1997) utilized similar methodology as the 
initial study; however, the primary purpose was to examine the age, gender, and velocity effects 
of walking on spinal kinematics. Participants performed overground walking at two, randomized 
velocities: self-selected and faster than self-selected. Participants were asked to walk at a normal, 
31 
 
comfortable pace for the former. During the latter condition, participants were instructed to adopt 
a gait pattern as if they were hurrying for an appointment. Height normalized velocity showed no 
difference between groups Senior Male (SM ? 50 years of age), Senior Female (SF ?50 years of 
age), Junior Male (JM < 50 years of age), Junior Female (JF < 50 years of age) at a normal, 
freely chosen velocity; however, during a faster condition there were significant differences in 
velocity between all groups except JM and JF. Instructions during the fast condition consisted of 
asking participants to choose a velocity which was comfortable, but similar to that which would 
be adopted if hurrying to a meeting. In addition, there were gender differences for height 
normalized step length during the free walking conditions, as was age, with seniors displaying 
significantly shorter strides. Similar trends were found for the faster condition as in the free 
walking condition; however, SF and SM displayed a significant gender effect with SM 
displaying a significantly longer height normalized stride length. With regard to pelvic and 
lumbar motion, similar motion patterns were exhibited by all participants at both speeds. During 
the free-walking condition, pelvis motion appeared to exhibit a period of stabilization, yet during 
the faster condition, the pelvis and trunk appeared in a constant state of motion. Overall, similar 
pelvic girdle rotations were observed between genders as in the earlier study, however, the 
amplitude of pelvic girdle movements increased with a slight offset in timing of events noted. 
Specifically, while positive pelvic girdle tilt still coincided with single limb support, posterior 
pelvic tilt began occurring prior to mid-stance. The authors also observed significant age effects 
on the amplitude of displacement for the thorax, lumbar spine, and pelvis; however the authors 
attributed this to reduced walking velocity of the SM and SF. This final conclusion of the authors 
was unusual given the lack of significant difference between height-normalized velocities for the 
free-walking conditions.  
32 
 
 The final study by Crosbie and Vachalathiti (1997) examined the synchronicity of the 
pelvic and hip kinematics during walking. As before, the authors utilized identical participant 
samples and similar methodology for data collection. Also included in this study were the two 
walking velocities previously described: self-selected and faster than self-selected. The authors 
found that males tended to demonstrate significantly greater vertical displacement at both 
velocities. Conversely, females demonstrated significantly greater pelvic rotation. Interestingly, 
the authors utilized separate multiple regressions to conclude that changes in step length 
contribute to larger changes in hip flexion amplitude in females than males. Furthermore, the 
authors assert that as walking velocity increased, a pseudo-locking of phases occurred, thereby 
concluding the pelvic contribution is of minimal consequence to normal gait. While noteworthy, 
more recent findings suggest this hypothesis is incorrect (Y. Huang et al., 2010; Y. P. Huang et 
al., 2011; C. J. Lamoth, Beek, et al., 2002; C. J. Lamoth, Meijer, et al., 2002; C. J. C. Lamoth et 
al., 2006).  
 In 1999, Taylor and colleagues examined the effect of walking velocity on pelvic and 
lumbar spine kinematics. The goal of this study was to better understand the effects of slower 
walking velocities on the spine and pelvis so as to better understand changes in pathological gait. 
Specifically, since reduced walking velocity is often witnessed in individuals with low back or 
pelvic girdle pain it was hoped to develop an understanding of lumbopelvic movements during 
slower walking speeds. During treadmill walking, PEAK 3-D motion measurement was utilized 
to record 27 participants that were randomly allocated to one of two walking velocities: self-
selected or 60% of self-selected overground walking velocity. During the trials, the speed of the 
treadmill was matched to the previously recorded overground walking velocity of that 
participant. The authors found a significantly lower pelvic list, or lateral pelvic girdle rotation 
33 
 
and transverse pelvic girdle rotation in the slow walking group, but no significant differences in 
pelvic tilt. The authors also found significantly lower lumbar lateral flexion relative to the pelvis 
in the slow walking group, yet significance was not reached for lumbar rotation or flexion-
extension, contradicting earlier research (Stokes, Andersson, & Forssberg, 1989). Interestingly, 
the authors noted that when the frame of reference was changed from the pelvis to a global 
reference, the significant differences for lumbar lateral flexion disappeared which was in 
agreement with earlier findings (Crosbie et al., 1997b; Krebs, Wong, Jevsevar, Riley, & Hodge, 
1992; Opila-Correia, 1990). Therefore, the authors asserted that spinal movements should be 
defined relative to the pelvis when studying underlying intervertebral movement during gait. 
 Lamoth, Beek, and Meijer (2002) published a similar paper examining thorax-pelvic 
coordination in the transverse plane during gait. Ten participants took part in the study, 
conducted with similar methodology to van Emmerick and Wagenaar (1996), however, unlike 
earlier studies, Lamoth and colleagues (2002) introduced the use of Discrete Fourier Transform 
to compare continuous relative phases. Rather than attempting to manipulate stride length or 
cadence, however, the authors only altered velocity. Beginning at 1.4 kmh-1, walking velocity 
was increased in 0.8 kmh-1 increments until a maximum of 5.4 kmh-1 was reached. In more 
traditional gait units, these values are equal to a beginning velocity of 0.39 ms-1 and ending 
velocity of 1.5 ms-1, with increases in intervals of 0.22 ms-1. The authors found that at lower and 
intermediate walking velocities, maximal thorax rotation occurred at 85 ? 95% of the gait cycle 
(prior to the secondary foot contact), yet almost always coincided with foot contact at higher 
walking velocities. Furthermore, the thorax-pelvis RP was relatively in-phase at low walking 
velocities, implying little counter-rotation, yet shifted to more anti-phase as walking velocity 
increased. Therefore, the results indicated pelvic-thorax coordination was significantly altered by 
34 
 
walking velocity, with increased spinal rotation at higher walking velocities; however, as the 
effects of stride and cadence were not examined, the application of these findings is limited.  
 In 2008, Bruijn and colleagues published findings examining the effect of pelvic and 
thorax transverse plane rotations to total body angular momentum. Similar to Lamoth et al. 
(2002), Bruijn and colleagues (2008) implemented a Discrete Fourier Transform to examined 
effects of walking velocity on segmental coordination during gait. Specifically, they recorded the 
kinematics of nine, healthy male participants walking on a treadmill at nine different velocities, 
ranging from 0.55 ms-1 to 1.44 ms-1. Like previous research, the authors found that the pelvis-
thorax coordination shifted from in-phase to out-of-phase as velocity increased. Interestingly, 
while the pelvis began to shift to a more in-phase state with the femur as velocity increased, the 
thorax became more out-of-phase with the femur at greater velocities. In determining the angular 
momentum contributions of individual segments to total body angular momentum, the authors 
reported that the pelvis and thorax only account for ~10% of total body angular momentum 
during walking. Conversely, the legs and arms account for almost 90% of the total body angular 
momentum. The authors therefore argued that, despite the conclusions put forth by Ducroquet 
and colleagues in 1968, pelvic-thorax coordination is relatively unimportant to the organization 
of total body angular momentum, indicating a need to develop a better understanding of the 
pelvic step (Bruijn et al., 2008).  
Most recently, Huang and colleagues (2010) published a paper examining the effects of 
stride length and stride frequency on trunk coordination during walking. Sixteen participants 
walked on an instrumented treadmill at predetermined velocities (0.5 ms-1, 1.0 ms-1, and 1.5 ms-
1) while wearing a custom made light emitting diode (LED) marker set attached at the T6 level, 
pelvis, and a single upper leg. While walking velocity was controlled by the authors, the 
35 
 
manipulation of stride length and cadence could be considered less than desirable. Likewise, 
information concerning footwear, or lack thereof, was not provided. Stride length, and thereby 
cadence, was manipulated by asking participants to walk a (1) self-selected step length, (2) larger 
than self-selected, and (3) smaller than self-selected. Descriptive statistics indicated that the 
author?s instructions were successful in achieving different stride lengths; however, there was a 
significant step x speed interaction, indicating the manipulation of stride length was less effective 
at higher velocities. Conversely, research on overground walking has indicated the use of an 
external metronome as an effective manner in which to manipulate cadence and velocity, thereby 
altering stride length (Romer et al., 2012; Thaut, Miltner, Lange, Hurt, & Hoemberg, 1999). As 
described by Huang and colleagues (2010), when the extremes of the movements are in-phase, or 
maximum left pelvic girdle rotation coincides with maximum left thorax rotation, the relative 
phase is 0? and is indicative of a synchronous movement of the segments. Conversely, when the 
extremes of the movements occur at opposite time points, the relative phase equals 180? and is 
indicative of an out-of-phase movement of segments. Despite the limitations of the former study, 
results indicated that with increasing step length, thorax-pelvis RP, thorax-leg RP, and rotational 
amplitudes of the thorax and pelvis became significantly larger. Similar to previous findings by 
Bruijn et al. (2008), pelvis-leg RP decreased with increasing step length. Huang and colleagues 
(2010) also found that there were significant increases in thorax-leg RP, thorax-pelvis RP, and 
spinal rotation as walking velocity increased. The results of the study suggest that during 
treadmill walking, increases in thorax-pelvis RP, decreased pelvis-leg RP, and the amplitude of 
pelvic, thorax, and spinal rotations increased as a result of increased stride length rather than 
cadence.  
36 
 
Because the product of stride length and cadence determine walking velocity, the 
aforementioned results, with an emphasis on those by Huang and colleagues (2010), provide an 
interesting basis for a hypothesis of changes in barefoot vs. shod gait. As barefoot locomotion is 
characterized by a reduced stride length and higher cadence relative to shod walking, it could be 
hypothesized that individuals will display more in-phase thorax-pelvic RP, decreased spinal 
rotation, and reduced rotational amplitudes during barefoot gait relative to shod gait. 
Furthermore, tactile sensitivity and feedback has been previously shown to be an important 
component in the management of gait kinematics; however, plantar sensation appears important 
in the managing of onset and magnitude of muscle activation (Bird et al., 2003; Chen, Nigg, 
Hulliger, & de Koning, 1995; Kelleher, Spence, Solomonidis, & Apatsidis, 2010; Kurz & 
Stergiou, 2003; Murley et al., 2009; Nurse et al., 2005; Nurse & Nigg, 2001; Scott, Murley, & 
Wickham, 2012). Though some researchers have found an increased physiological and 
mechanical economy during shod gait, the primary foci of these papers have been on changes in 
the lower leg (Divert et al., 2005; Divert et al., 2008; Lieberman et al., 2010). Considering the 
semi-rigid linkage that exists between the distal lower appendicular skeleton and the proximal 
lower appendicular and axial skeletons, it would seem imperative to understand the effects of 
footwear and plantar sensation on trunk and pelvic movements during gait.  
In summary, research supports considerable differences in the RP of healthy individuals 
as velocity increases. Specifically, the thorax-pelvis and thorax-leg appear to rotate in a more 
asynchronous manner as velocity increases. Similarly, amplitude of segmental rotation (i.e. 
thorax, pelvis, and spine) appears to increase as velocity increases. Conversely, the thorax-leg 
RP has previously been reported to become more synchronous as velocity increases (Bruijn et 
al., 2008; Y. Huang et al., 2010). Minimal information is available concerning the independent 
37 
 
variables that produce velocity (i.e. stride length, cadence), though research suggests that 
increased step length may result in more asynchronous movements and greater rotational 
amplitudes (Y. Huang et al., 2010; Y. P. Huang et al., 2011). While footwear has been shown to 
significantly alter lower leg gait mechanics, currently unknown is the extent, if any, of the 
footwear effect on trunk coordination during gait. Research examining phase differences in low 
back or pelvic girdle pain populations suggest that these individuals tend to self-select segmental 
movements that result in more synchronous thorax-pelvis RP, indicative of a tenuous link 
between afferent feedback in gait mechanics and relative phase changes (Y. Huang et al., 2010; 
C. J. Lamoth, Meijer, et al., 2002; C. J. C. Lamoth et al., 2006; W. Wu et al., 2002; W. Wu et al., 
2004; W. H. Wu et al., 2008). The self-selected velocities of pathological populations tend to be 
lower than the healthy counterparts; however, at comparable velocities, symptomatic individuals 
tend to display a higher cadence, or greater rotational amplitudes for a given stride length, 
thereby decreasing the relative phase differences of the thorax and pelvis. Furthermore, RP 
changes in pathological populations are suggestive of manners in which spatiotemporal variables 
may be manipulated when the natural constraints of the musculoskeletal system are altered. 
Remaining to be answered though, is the type of adaptations that arise, if any, by healthy 
individuals in response to the reduced plantar sensation caused by footwear. The present project 
intends to examine the effect of the independent manipulation of velocity and cadence on RP 
changes to better understand mechanisms responsible for altering RP. Therefore, the present 
project will attempt to differentiate the effect of footwear and augmented plantar feedback on 
changes in segmental coordination and muscular activation during walking.  
 
 
38 
 
Section 3: Influence of Footwear on Human Gait 
 While the author is unaware of any studies examining the effect of footwear on the 
coordination of the thorax and pelvis during gait, there is a wealth of information concerning the 
effect of footwear on human gait. This includes minimalist footwear, barefoot locomotion, 
sandals, flip-flops, varying height boots, athletic footwear, and numerous types of casual 
footwear. In fact, this is still a rapidly expanding field of research that regularly garners clinical 
and industrial interest. A complete review of all the various types of footwear would be beyond 
the scope of the current project. Instead the focus of this section will be on the differences 
between barefoot locomotion and shod gait. While the present project intends to focus on 
differences in gait mechanics during walking, the vast majority of research has focused on 
differences in running mechanics. Nonetheless, a pool of research is available to examine 
differences in walking mechanics. Likewise, the similarity between running and walking does 
allow for one to draw some global conclusions from differences in barefoot and shod during both 
types of gait.  
 While the examination of differences in barefoot and shod gait mechanics seems like a 
relatively elementary comparison, it has recently experienced a considerable upsurge in research 
and pop-culture interest. This has been partially attributable to the 2009 book by Christopher 
McDougall, which among other items, discussed the relative dominance in ultra-distance running 
events from the Native American Tarahumara tribe. A key component of their lifestyle is the 
relatively minimalist footwear that is used by the tribe members in everyday and athletic life, 
resembling a sandal with a thin piece of leather material serving as the sole of the shoe. At 
approximately the same time, Daniel Lieberman, a researcher and professor at Harvard 
University, was in the process of publishing a series examining the evolutionary role of the 
39 
 
human foot (Bramble & Lieberman, 2004; Lieberman et al., 2010). In 2010, Lieberman 
published a paper examining the gait mechanics of habitually shod versus barefoot runners that 
received considerable notoriety.  
 Lieberman and colleagues (2010) found that habitually shod runners exhibited markedly 
different gait mechanics during running. Specifically, the authors found that habituated shod 
runners experienced a marked heel stride while habituated barefoot runners displayed a forefoot 
strike. The former is characterized by an initial and pronounced heel contact, with the foot later 
transitioning towards footflat. The latter has been defined as an initial contact of the forefoot, 
followed by ground contact of the heel, resulting in a flat foot. Mechanically, forefoot striking 
has been found to be a more economical method of running (Divert et al., 2005; Lieberman et al., 
2010). Though the mechanism is not completely understood, it is believed that it allows the 
intrinsic and extrinsic musculature of the foot and lower leg to more successfully utilize the 
stretch-shortening cycle during a rapid, powerful activity such as running. Forefoot strikers also 
typically exhibit a decreased stride length and increased cadence during equivalent velocities. 
Lieberman et al. (2010) noted similar differences between the habitual shod and habitual 
barefoot runners. Furthermore, the authors noted that barefoot runners displayed significantly 
greater hip extension and plantarflexion at initial foot contact. Despite the absence of the 
cushioning effects of footwear, barefoot runners displayed a significantly reduced impact 
transient (e.g. rate of loading) as compared to heel strikers.  
 Kristen et al. (1998) reported similar changes in loading patterns as Lieberman et al. 
(2010), however, the changes were observed in children during walking. The authors examined 
30 young children (< 5 years) during barefoot gait and with three different types of shoes. Using 
3D kinematic and dual ground reaction force platforms, the authors found that the participants 
40 
 
displayed a significant increase in the ground contact time (velocity was not reported), shifting of 
maximal load weight from the rear foot towards mid-foot, and a cushioning effect during initial 
impact of ground contact. The authors concluded that footwear does significantly alter loading 
patterns of gait, even for young children (Kristen et al., 1998).  
 In an attempt to understand differences in foot motion during locomotion, Morio and 
colleagues (2009) used three conditions (i.e. barefoot, and sandals with two levels of midsole 
hardness). Participants performed a series of overground walking and running trials for all three 
conditions. To control for any footwear effects on either walking or running velocities, gait speed 
was held constant through the use of a photocell timing gate (Morio, Lake, Gueguen, Rao, & 
Baly, 2009). The authors reported significant footwear effects, with barefoot locomotion 
displaying significantly greater forefoot eversion. Furthermore, the rate of forefoot eversion also 
occurred at a significantly higher rate than during shod locomotion. The authors also reported 
that the sole of the sandals constrained the torsional and adduction range of motion of the foot, 
restricting the natural motion of the foot and toe-off strategy. Though not examined in this study, 
the authors argued that the kinetics of the push off phase would likely be considerably altered 
due to the imposition of the motion patterns on the foot.  
 Braunstein et al. (2010) presented another study that examined the effects of footwear on 
running mechanics. The authors used five different types of running shoes, plus an additional 
barefoot condition, to compare the gear ratio during running on grass. The gear ratio is defined as 
the ratio of the moment arm of the ground reaction force to the moment arm of the counteracting 
musculotendinous unit, an indication of joint loading and mechanical efficiency. Fourteen, 
endurance trained runners took part in the study which consisted of running along a 20m  x 1m 
running track at a predefined velocity of 4.0 ? 0.2 ms-1. A Kistler ground reaction force plate was 
41 
 
mounted flushed with the ground. The authors reported no significant footwear differences in 
gear ratios or lower extremity moments for any of the five different shoes, however, results 
indicated that wearing running shoes significantly altered the gear ratios for the ankle and knee 
joints. Specifically, barefoot running was associated with an increased gear ratio for the ankle 
during early stance, yet a lower ratio during late stance. The knee gear ratio was lower during 
midstance in comparison to the shod conditions. The authors concluded that footwear results in a 
higher mechanical stress on the knee during the stance phase, but an increased mechanical 
advantage of the plantarflexors during toe-off. 
 Following the work by Kristen et al. (1998), a 2011 meta-analysis by Wegener and 
colleagues performed a systematic review of the biomechanical effects of shoes on children 
during walking and running. The inclusion of a barefoot and at least one shod condition were 
required for inclusion in the screening process. A total of 11 studies were included, with a 
sample size ranging from 4 ? 498 and a Mean Quality Index of 20/32. The results of the meta-
analyses found that relative to barefoot gait, shod walking increased a number of spatiotemporal 
variables including velocity, step length, step time, base of support, double-support time, stance 
time, and time to toe-off.  Furthermore, considerable changes were determined in joint 
kinematics with increases in sagittal plane rearfoot range of motion (ROM), maximum ankle 
plantarflexion, ankle ROM, subtalar ROM, foot lift to maximum plantarflexion, subtalar ROM, 
knee ROM, and tibialis anterior activity. Similar findings have been reported in adults during 
barefoot and shod locomotion (Divert et al., 2005; Divert et al., 2008; Kelleher et al., 2010; 
Lieberman et al., 2010; Nurse et al., 2005). Therefore, it appears that children and adults tend to 
walk faster, take longer steps, and have increased lower extremity ROM and muscle activity 
when wearing shoes. A compounding factor that has not adequately been investigated is the 
42 
 
interaction between velocity and muscle activity. While increased tibialis anterior activity may 
have been associated with changes in weight to the distal segment of the foot or alterations in 
tactile feedback, it could also be associated with the increased velocity associated with shod 
walking.  
 In an attempt to answer the question of whether changes in spatiotemporal gait 
parameters were mechanical in origin, Divert and colleagues (2008) designed a study using ultra-
thin diving socks that could be externally loaded with additional mass. The authors recruited 
twelve, recreationally trained runners for participation in the study which included six, 4-minute 
running bouts at 3.61 ms-1. During each trial, participants wore only diving socks (unloaded, 
loaded to 150g, and loaded to 350g), traditional running shoes (loaded to 150g or 350g) or were 
completely barefoot. Oxygen consumption and three-dimensional ground reaction forces were 
collected through the entirety of each trial. The authors found a significantly higher stride 
frequency, anterior-posterior impulse, vertical stiffness, leg stiffness, and mechanical work 
during the barefoot conditions relative to either shod condition. Furthermore, net efficiency (e.g. 
metabolic and mechanical components) decreased during the shod conditions relative to the 
barefoot conditions. Based on the results of this study, the authors concluded that the main role 
of the shoe is to attenuation foot-ground impact by adding damping material; however the 
effectiveness of footwear in actually performing this task has previously been called into 
question with Robbins and Gouw (1990) reporting no change in shock attenuation during 
running at 3.89 m*s-1 on a treadmill. Furthermore, an earlier study found no significant 
differences in impact forces between standard running shoes and running shoes that contained a 
50% increase in heel cushioning (Clarke, Frederick, & Cooper, 1983). Nevertheless, it was 
43 
 
hypothesized that changes in the response of the elastic and passive tissues in footwear could be 
an explanatory reason for reductions in net efficiency (Divert et al., 2008).  
 In conclusion, footwear has been found to significantly alter the kinematics and kinetics 
of walking and running gait. While velocities tend to be fairly similar between barefoot and shod 
locomotion, barefoot gait is associated with a reduced stride length and higher cadence than shod 
gait (Divert et al., 2008; Larsen, Weidich, & Leboeuf-Yde, 2002; Lieberman et al., 2010; Morio 
et al., 2009; Wegener et al., 2011). As a result, differences in joint kinematics have been 
reported, with barefoot walkers exhibiting increased plantarflexion and decreased hip flexion at 
initial contact (Kristen et al., 1998; Lieberman et al., 2010; Wegener et al., 2011). Shod gait has 
been associated with important kinetic differences including an increased impact transient, 
decreased vertical and leg stiffness, and increased anteroposterior ground reaction forces. 
Furthermore, Braunstein et al. (2010) reported significant differences in the joint kinetics and 
mechanical efficiency in barefoot and shod locomotion. Therefore, though significant alterations 
in lower extremity gait mechanics have been previously reported in the literature, the author is 
unaware of any existing research that have examined the footwear effect on relative phase 
differences during gait. Despite the significant differences that have been found in lower 
extremity kinetics and kinematics, peer-reviewed information regarding the footwear effect on 
pelvic and low back mechanics is relatively sparse. The present project hopes to fill this void by 
examining the effect of footwear on relative phase changes and pelvic girdle mechanics during 
walking. Additionally, the present project will examine the efficacy of augmented plantar 
feedback in altering gait mechanics. A secondary goal of the augmented plantar feedback is to 
determine if enhanced plantar sensation will cause gait mechanics to shift towards a more natural 
44 
 
form of locomotion as has previously been hypothesized by some researchers (Bird et al., 2003; 
Kelleher et al., 2010; Kurz & Stergiou, 2003).  
Section 4: Role of Plantar Sensation and Tactile Feedback in Human Locomotion 
In addition to the aforementioned mechanical influences of footwear, there also appears 
to be a considerable interaction between footwear and gait on the onset and magnitude of muscle 
activity. Especially interesting are changes in the onset of muscle activity, which may have 
implications for the loading of active and passive tissues. While the predominant research 
investigating muscular activation has focused on the lower extremity, there may be important 
implications for the lumbopelvic musculature as previous research has shown a relationship 
between lower extremity and core musculature recruitment (Allet et al., 2009; Bird et al., 2003; 
Kersting, Janshen, Bohm, Morey-Klapsing, & Bruggemann, 2005; Li & Hong, 2007; 
Mundermann et al., 2001; Murley et al., 2009; Nurse et al., 2005; Wakeling et al., 2003). 
Delayed activation or altered recruitment of the core musculature may alter the pelvic girdle 
positioning during loading, thereby affecting the mechanical demands of the passive tissues 
(Willson et al., 2005). While there has not been extensive research in the area regarding footwear 
and muscle activation, it has become a recent subfield of this expanding area of research. 
Furthermore, research investigating muscular activation differences in individuals with reduced 
tactile sensitivity (e.g. pathological, mechanically, or chemically induced peripheral neuropathy) 
may provide a window into the effects of decreased sensory feedback on gait mechanics.   
 Peripheral neuropathy is often found in individuals in advanced stages of diabetes, a 
metabolic condition that may compound any deduction from findings to healthy populations. 
However, when controlling for anthropometric factors, researchers have found considerable 
differences in healthy controls, diabetic individuals without neuropathy, and diabetic individuals 
45 
 
with neuropathy (PN). For instance, Allet and colleagues (2009) found that individuals with PN 
displayed significant differences in a number of kinematic variables (i.e. velocity, cadence, 
single and double support percentage, cycle time, stride length, and hip angle) compared to 
healthy individuals. Furthermore, individuals with PN also displayed a significantly higher 
stride-to-stride variability as compared to healthy individuals, which may be a sign of decreased 
postural control and increased core instability. Similarly, other researchers have found that 
individuals with PN exhibited significantly greater center-of-pressure trajectory during a 
standard balance test (90 seconds, normal stance, eyes closed) as compared to healthy controls 
(Manor & Li, 2009). Interestingly, other researchers have found that augmented plantar 
facilitation may increase postural control in populations with balance deficits. For instance, Maki 
et al. (1999) found a significant decrease in the incidence of extra limb movements and center-
of-pressure trajectory in response to postural perturbations when older (65-73 years of age) and 
younger (23-31 years of age) individuals wore footwear that provided augmented plantar 
facilitation. Other research has found that providing healthy subjects with augmented plantar 
feedback resulted in significant changes in plantar pressure distribution during walking (Chen et 
al., 1995). Specifically, Chen et al. (1995) reported an increased midfoot pressure and decreased 
pressure in the toe area when participants were provided increased plantar sensory input. From a 
mechanical standpoint, other research has shown an altered windlass mechanism in individuals 
with PN, with a greater vertical GRF impulse and impact transient in individuals with PN as 
compared to healthy controls (D'Ambrogi, Giacomozzi, Macellari, & Uccioli, 2005; Maki et al., 
1999). Therefore, it appears that peripheral sensory feedback may not only be an important factor 
in lower extremity kinematics during gait, but tactile feedback may be an important modulator in 
thorax and pelvic movements and stability.   
46 
 
 A drawback of the aforementioned research is the lack of an investigation into changes in 
efferent activation as a result of changes in afferent feedback. Conversely, Savelberg and 
colleagues (2010) measured the electromyography (EMG) of seven lower extremity muscles and 
found that, regardless of walking velocity, individuals with PN exhibited a significant decrease 
in the duration of dorsiflexor and knee extensors musculature activity during the stance phase of 
gait. These results are especially intriguing as there were no reported differences in the joint 
kinematics or spatiotemporal variables of the pathological populations relative to the healthy 
controls (Savelberg et al., 2010). While Savelberg et al. (2010) reported no changes in the onset 
of muscle activity others have investigated the effect of plantar facilitation on changes in 
musculature timing during the gait cycle. For example, a 2003 study investigated the effect of 
various forefoot wedge and heel lift designs on the activity of the erector spinae (ES) and gluteus 
medius (GMed) in healthy individuals (Bird et al., 2003). Interestingly, bilateral heel lifts and all 
three forefoot wedge designs caused a significant delay in the onset of GMed activity while the 
heel lifts caused a significantly earlier onset of ES muscle activity. The delay in GMed muscle 
activity accounted for approximately 2% of the gait cycle while the decrease in the onset of ES 
muscle activity accounted for approximately 4% of the gait cycle. Furthermore, despite the 
significant alteration in onset of muscle activity, Bird and colleagues (2003) reported no 
significant changes in the amplitude of muscle activity. Similarly, Murley et al. (2009) reported 
significant changes in lower back and lower limb muscle activity in response to an elevated heel 
height, with the alterations becoming more pronounced as heel height increased. Interestingly, it 
has also been found that habitual wearers of elevated heel heights tend to display greater 
fatigability of the peroneus longus and the lateral gastrocnemius (Gefen et al., 2002). 
Unfortunately, most athletic footwear contains an elevated heel similar in magnitude to those 
47 
 
utilized by Bird et al (2003); however, researchers have been unable to quantify the degree of the 
toe-heel height differential that will alter physiological and mechanical factors in movement 
(Chiu & Wang, 2007; Gross et al., 1991; Kerrigan, Todd, & Riley, 1998; Kersting et al., 2005; 
Larsen et al., 2002; Murley et al., 2009). Finally, it has also been reported that many types of 
orthotics cause a significant increase in the magnitude of the peroneal and anterior tibialis 
activity, while also increasing the duration of anterior tibialis activity during gait (Murley et al., 
2009).  It was recently reported that the amplitude and time to peak amplitude for the tibialis 
anterior, peroneus longus, and medial gastrocnemius were significantly different between 
barefoot and shod gait, with few differences reported between standard shoe designs (flexible 
shoe, stability running shoe) (Scott et al., 2012). Specifically, peak tibialis anterior and medial 
gastrocnemius amplitude was significantly larger, with an earlier occurrence found for the 
tibialis anterior, during shod gait as compared to barefoot gait. Conversely, peak peroneal 
amplitude was reduced during shod walking as compared to barefoot walking. Therefore, in 
addition to the importance of peripheral sensory information in managing lower extremity, 
pelvic, and torso kinematics, it appears as though footwear and footwear components are capable 
of producing significant changes in the muscle activity of the aforementioned body segments.  
  Wakeling and colleagues (2003) investigated whether muscle activity in the lower 
extremity is used to dampen soft-tissue resonance during gait, which displays considerable peaks 
at foot contact. By comparing the force transfer from the ground to the soft tissues, as well as the 
muscle activity of several lower extremity muscles, the authors found that increases in biceps 
femoris and lateral gastrocnemius activity corresponded to an increased vibration dampening. 
Therefore, the authors concluded that the evidence supported the hypothesis that muscle activity 
dampens soft-tissue resonance at foot strike, a potentially important finding for understanding 
48 
 
the loading of passive tissues in the lower extremity and low back. Furthermore, the authors 
assert that shoe design may be used to modify muscle activity and joint loading during walking 
and running (Wakeling & Liphardt, 2006; Wakeling et al., 2003). To test this hypothesis, Nurse 
and colleagues (2005) performed a study to determine the effect of textured and smooth inserts 
on lower extremity muscle activity, limb kinematics, and joint kinetics. Amazingly, the authors 
found that the textured insert resulted in a significant reduction in the total energy required by the 
soleus and tibialis anterior during the stance phase, with a reduction in rectus femoris activity 
approaching, but not reaching, statistical significance. Furthermore, there was also a significant 
increase in the ankle angle, indicating greater plantarflexion, at heel strike when using the 
textured insert, indicating the ability of footwear to alter joint kinematics (Nurse et al., 2005). 
Finally, when wearing the textured insert was compared to the smooth insert, the participants 
also displayed a significant increase in the time to impact peak, implying a reduced impact 
transient, and a slight, but significant increase in the active peak force during the first half of the 
stance phase. 
 In summary, there is considerable support for the ability of footwear to significantly alter 
the onset and magnitudes of lower extremity muscle activity, with some research finding 
significant alterations in pelvic and trunk musculature. Not only would altered onset of muscle 
activation have considerable effects on joint and segment positioning, as emphasized by the 
kinetic chain model of movement, but would also considerably alter the loading of the passive 
tissues (McMullen & Uhl, 2000). As suggested by Wakeling and colleagues (2003), the active 
tissues of the body are important in the maintenance of vibrational and soft-tissue resonance 
occurring as a result of ground contact. Though caution should be considered in its application, 
individuals with peripheral neuropathy have shown altered gait kinematics (i.e. velocity, 
49 
 
cadence, single and double support percentage, cycle time, stride length, and hip angle), and 
have been hypothesized to have a reduced postural control, as compared to healthy controls 
(Allet et al., 2009). Similarly, other researchers have found that individuals provided with 
augmented plantar feedback displayed significantly improved postural stability during 
unpredictable postural perturbations (Maki et al., 1999). Together, the literature base is 
suggestive of the ability of augmented plantar facilitation to significantly alter lower extremity 
and torso kinematics and muscular activation during gait; however minimal information exists in 
this area, with past research focusing on special populations, significant footwear alterations (e.g. 
heel lifts), or isolated to the lower leg differences. The present project examined the effect of 
augmented plantar facilitation on gait kinematics and the muscular activation of torso and lower 
extremity at key gait cycle events (e.g. initial contact, single support). Furthermore, the present 
project investigated differences in gait kinematics and muscular activation between barefoot, 
augmented, and shod conditions, an area that the author believes is a relatively novel concept to 
the gait research community. It is believed that the study approach provided a better 
understanding of the multi-dimensional link (e.g. physiological, mechanical, and neural) 
responsible for human movement.  
Section 5: Summary of the Literature 
Human locomotion is one of the more researched topics in biomechanics literature. 
During the initial stages of gait, the vast majority of the literature relied on descriptive results to 
explain the general movement patterns during walking. For example, Saunders et al. (1953) 
provided an early basis of pelvic and lower extremity movements during gait, collectively 
referred to as the Six Determinants of Gait. Similarly, Ducroquet and colleagues (1968) were one 
of the first groups to discuss the relative relationship between trunk, pelvic, and leg movements 
50 
 
during gait. Recently, more quantitative research has been performed to understand relative 
phase changes as a result of changing spatiotemporal variables (Crosbie & Vachalathiti, 1997; 
Crosbie et al., 1997a, 1997b; Y. Huang et al., 2010; Y. P. Huang et al., 2011; C. J. Lamoth, 
Beek, et al., 2002; C. J. Lamoth, Meijer, et al., 2002; C. J. C. Lamoth et al., 2006; Taylor, Evans, 
& Goldie, 2003; Taylor et al., 1999; van Emmerik & Wagenaar, 1996; W. Wu et al., 2002; W. 
Wu et al., 2004; W. H. Wu et al., 2008). Results by Huang and colleagues (2010) suggest that, 
during treadmill walking, increases in thorax-pelvis RP, decreased pelvis-leg RP, and the 
amplitude of pelvic, thorax, and spinal rotations increased as a result of increased stride length 
rather than cadence. As barefoot locomotion is characterized by a reduced stride length and 
higher cadence relative to shod walking, it could be hypothesized that individuals will display 
more in-phase thorax-pelvic RP, decreased spinal rotation, and reduced rotational amplitudes 
during barefoot gait relative to shod gait. 
Considerable research effort has been directed at understanding the effects of footwear on 
spatiotemporal gait parameters. While the vast majority of the research has been spent 
investigating  various types of footwear, an emerging area of focus is on understanding the 
underlying factor for differences in barefoot locomotion. Bird and colleagues (2003) reported 
that bilateral heel lifts and three forefoot wedge designs caused a significant delay in the onset of 
GMed activity while the heel lifts caused a significantly earlier onset of ES muscle activity. 
Similarly, Murley et al. (2009) reported significant changes in lower back and lower limb muscle 
activity in response to an elevated heel height, with the alterations becoming more pronounced as 
heel height increased.  
While the aforementioned authors significantly altered the shoe design, Nurse et al. 
(2005) took a relatively more passive approach. Specifically, when comparing a textured insole 
51 
 
and smooth insole, the authors found that the textured insert resulted in a reduction in the total 
energy of three lower extremity muscles (tibialis anterior, soleus, rectus femoris). Furthermore, 
the authors also found a significant increase in the ankle angle, with greater plantarflexion at heel 
strike when using the textured insert. The importance of tactile sensitivity appears to be an 
important factor in the modulation of gait mechanics during walking and running. As suggested 
by Kurz and Stergiou (2003), the reduced plantar sensation and kinematic variability associated 
with shod locomotion may be a basis for the etiology of many lower extremity injuries. 
Furthermore, afferent feedback during gait appears to be an important factor in the efferent 
response of lower extremity muscle activity during gait, altering the loading patterns of active 
and passive tissues (Bird et al., 2003; Chen et al., 1995; Kelleher et al., 2010; Maki et al., 1999; 
Murley et al., 2009; Wakeling & Liphardt, 2006; Wakeling et al., 2003).  
The current project examined the effects of changes in segmental coordination, gait 
kinematics, and muscle activation in response to altered spatiotemporal variables. Another 
purpose of the current study was to examine the effect of footwear on segmental coordination, 
gait kinematics, and muscle activation. Through this research, an expanded understanding of the 
effects of footwear on tissue loading during a high volume, cyclic activity such as walking has 
been developed. The methodology implemented in the present study provided a unique and 
comprehensive approach aimed at understanding changes in active tissue response and loading of 
the musculoskeletal system during gait.  
 
 
 
 
52 
 
Chapter III. Methodology 
Walking is the most common form of locomotion for humans, and despite its relative 
simplicity, relies on the delicate coordination of numerous joints, muscles, and body segments. 
The pelvis contains a large number of muscles which are responsible for providing the stability 
from which movements can be initiated and a base from which propulsive forces can be 
generated. Likewise, the lumbopelvic complex is responsible for numerous important 
movements in gait, as identified in descriptive ways by Saunders and colleagues (1953) in the 
Six Determinants of Gait and Ducroquet et al. (1968) with the pelvic step. In addition to these 
early descriptive studies, researchers have been attempting to quantify and accurately describe 
the relationship between thorax, lumbar, pelvic, and lower extremity movements during gait. 
Despite these efforts, the patterns and mechanisms responsible remain partially elusive, 
demanding a unique and innovative approach to understand the movements of the lumbopelvic 
complex during gait. Furthermore, footwear is a necessary component of modern life yet has 
been shown to significantly alter lower extremity kinematics, kinetics, and muscular activity. 
Despite these findings, there is minimal information concerning the effect of footwear on the 
mechanics and muscle activity of the lumbopelvic complex. Therefore, the purpose of this study 
was to better understand the effects of altered spatiotemporal variables and footwear on relative 
phase differences and muscle activation during walking. Specifically, this study investigated : (1) 
the effects of velocity and cadence on lumbopelvic rhythm during overground walking; (2) the 
effects of footwear on lumbopelvic rhythm during gait; (3) the effects of enhanced tactile 
feedback (i.e. barefoot, textured insole) on the onset of muscle activity at foot strike; (4) the 
effects of spatiotemporal and footwear variation on the magnitude of lumbar (i.e. erector spinae) 
and lower extremity muscular activity (e.g. gluteus medius, biceps femoris, soleus, peroneus 
53 
 
longus) during the stance phase of gait, and (5) differences in the kinematics and 
electromyography between males and females. The following chapter presents the methodology 
that will be used to address the purposes of the present study and includes the following sections: 
(a) participants, (b) setting, (c) instrumentation, and (d) experimental design and analysis. The 
research protocol was submitted, reviewed, and approved by the Auburn University Institutional 
Review Board for Research Involving Human Subjects.  
Participants 
 Fifty-five college-aged males (n=27) and females (n=28) between the ages of 19-35 years 
were recruited as participants for the study. A power analysis was conducted (effect size = 0.25, 
alpha = 0.05, and power = 0.80) with G*Power v3.0.10 for Windows, determining that 40 
participants would be required to demonstrate significance (Faul, Erdfelder, Lang, & Buchner, 
2007). Therefore, 55 participants were chosen for inclusion to ensure adequate power. All 
participants wore identical clothing consisting of a compression tank-top and compression shorts. 
Prior to the beginning of data collection, participants were sized for footwear and were provided 
footwear for use throughout the duration of the protocol. Each participant completed a health 
screening survey and was excluded if an affirmative response was provided to any question that 
was deemed by the primary investigator to influence the participant?s performance in the study 
(Appendix A). Exclusion criteria included: (1) any current or recent injury (<1 year) to the lower 
extremity, pelvis, low back, or trunk, (2) any previous injury/illness that could jeopardize the 
successful performance of the requisite tasks, (3) any allergies to adhesives or adhesive type 
products, (4) any known balance deficits or inner ear disturbances. Participants indicated their 
willingness to participate by signing an Institutional Review Board approved Informed Consent 
document before data collection commenced (Appendix B).  
54 
 
Setting 
 All testing and data collection occurred within the Lower Extremity Laboratory at 
Auburn University (003 Kinesiology). The Lower Extremity Laboratory is a large enclosed 
laboratory with the necessary equipment to carry out this research. 
Instrumentation 
Kinematics 
 A 10 camera Vicon? MX motion analysis system (Vicon?, Los Angeles, CA, USA) was 
utilized to collect three dimensional kinematics of the lower extremity, lumbopelvic complex, 
thorax, and upper extremity during each condition. Ten cameras were positioned to record all 
three cardinal planes within a capture volume large enough to capture one full gait cycle (Figure 
9). Furthermore, the sampling frequency of the system was 200 Hz. Spatial positions of body 
segments were tracked through the use of 42 spherical, retroreflective markers at key anatomical 
landmarks. The markers used were 12mm in diameter (MKR-6.4, B&L Engineering, Tustin, CA, 
USA) and attached with double-sided tape (Double Sided Duck ?, ShurTech Brands, Avon, OH, 
USA). Markers were placed on the participant?s clothing and footwear or, when appropriate, 
directly on the skin. A modified Plug-In-Gait marker placement protocol was utilized. The 
modified marker protocol included the addition of medial markers on the elbow, knee, and ankle 
as well as markers placed bilaterally on the greater trochanter. Table 2 details marker 
descriptions and applied segments. Euler angles were calculated for the thorax, pelvis and right 
leg by The MotionMonitor (Innovative Sports Training, Chicago, IL, USA). Rotation sequences 
for the pelvis were Z Y? X?? while the thorax and right leg used an X Z? Y?? rotation sequence. 
55 
 
 Figure 9. Vicon motion capture camera orientation.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
56 
 
Table 2                 
Names and positions of markers used in Modified Plug-In-Gait Model.       
Marker(s) Name Position         Segment   
R/LFHD   Right/ Left Front of Head     Head   
R/LBHD   Right/ Left Back of Head     Head   
CLAV   Sternoclavicular Joint    Torso   
STRN   Inferior Sternum       Torso   
C7   Seventh Cervical Vertebrae     Spine   
T10   Tenth Thoracic Vertebrae     Spine   
R/SJO   Right/ Left Acromioclavicular Joint   Shoulder   
R/L LATELB Right/ Left Lateral Aspect of the Humeroulnar Joint Elbow   
R/L MEDELB Right/ Left Medial Aspect of the Humeroulnar Joint Elbow   
R/LWRA   Right/ Left Radius       Wrist   
R/LWRB   Right/ Left Distal End of Ulna     Wrist   
R/LFIN   Right/ Left Third Metacarpophalangeal Joint   Knuckle/ Finger   
R/LASI   Right/ Left Anterior Superior Iliac Spine   Pelvis   
R/LPSI   Right/ Left Posterior Superior Iliac Spine   Pelvis   
R/LTRO   Right/ Left Greater Trochanter     Leg   
R/LTHI   Right/ Left Lateral Aspect of the Femur   Upper Leg   
R/L LATKNE Right/ Left Lateral Tibiofemoral Joint     Knee   
R/L MEDKNE Right/ Left Medial Tibiofemoral Joint     Knee   
R/LTIB   Right/ Left Lateral Aspect of the Fibula   Lower Leg   
R/L LATANK Right/ Left Lateral Malleolus     Ankle   
R/L MEDANK Right/ Left Medial Malleolus     Ankle   
R/LTOE   Right/ Left Distal End of Second Metatarsal   Toe/ Foot   
R/LHEE   Right/ Left Calcaneus   Heel   
 
 The gait cycle for the current study was defined as beginning at initial contact of the right 
foot and ending at second initial contact of the ipsilateral limb. Other events included in the gait 
cycle were stance phase and swing phase (Figure 1). Furthermore, the cycle ensured that the 
entire body was within the capture volume for a full stride length, allowing the evaluation of 
specific points of interest. 
 
57 
 
Spatiotemporal Variables 
 In addition to the kinematic variables collected through the use of motion capture, 
spatiotemporal variables were collected through the use of an instrumented walkway (GAITRite, 
CIR Systems Inc., Havertown, PA, USA) (Figure 10). The 4.25m GAITRite mat was placed in 
the center of the capture volume, with a 3.5m lead-in and 3.5m walkout, for a minimum distance 
of 11.25m per trial. The pressurized mat collected footfall information at 100 Hz, providing real-
time feedback of the velocity and cadence of the participant.  
 
Figure 10. GAITRite Mat. 
 
Surface Electromyography (sEMG) 
 Thorax (right erector spinae) and lower extremity (bilateral gluteus medius, right biceps 
femoris, right soleus, and right peroneus longus) muscle activity were collected using six pairs of 
bipolar Ag-AgCL surface electrodes (Red Dot, 3M. St. Paul, MN, USA). Pre-amplified sEMG 
leads were connected to a Noraxon Telemyo 2400T V2 wireless transmitter (Noraxon USA Inc., 
58 
 
Scottsdale, AZ, USA) (Figure 11) which was connected to the Noraxon Telemyo 2400R ? World 
Wide Telemetry receiver (Noraxon USA Inc., Scottsdale, AZ, USA) (Figure 12) which recorded 
at a sampling rate of 1500Hz.  
 
 
Figure 11.  Noraxon Telemyo 2400T V2 wireless transmitter. 
 
 
 
Figure 12.  Noraxon Telemyo 2400R World Wide Telemetry receiver. 
 
Raw EMG signal was amplified with an input impedance of 10 M? and a set at 1000X 
(Knight & Weimar, 2011a). The data reduction procedures for the linear envelope detection of 
EMG occurred post-collection, including a band-pass filter to remove low-frequency and 
ambient noise (De Luca, 1997; Kamen & Gabriel, 2010; Knight & Weimar, 2011a, 2011b). The 
59 
 
signal was full-wave rectified and root mean squared at 50ms, which involved taking the 
absolute value of the raw EMG signal, with the high-frequency components being removed due 
to the filtering (De Luca, 1997; Kamen & Gabriel, 2010; Winter, 2009). Finally, the double 
threshold method was applied to linear enveloped data to determine onset of muscle activity. 
The double threshold method was employed to ensure maximum confidence in onset of 
muscle activity (De Luca, 1997; Kamen & Gabriel, 2010; Murley et al., 2009). This approach 
acts as a moving window on the linear envelope detected EMG. The double threshold method 
begins with calculating the mean and standard deviation of minimal EMG activity (e.g. absence 
of meaningful muscle contraction). It is recommended that the confidence intervals then be 
extended to 95% (??1.96?) or greater if baseline activity is significant (De Luca, 1997; Kamen 
& Gabriel, 2010). The corresponding level then signified the initial threshold value needed to 
signal onset of muscle activity. The second threshold required that once the amplitude of EMG 
activity reached the initial threshold value, it must stay above the threshold for some critical time 
period (De Luca, 1997; Kamen & Gabriel, 2010). While, the length of the window for baseline 
activity has been reported as not overly important to the validity of the final results, the length of 
the second window is highly dependent on the movement or task being studied.  Reported 
timeframes for the second criterion range from 10ms ? 50ms, with higher baseline activities 
associated with longer criterion windows (Kamen & Gabriel, 2010). In the present study, a 
window of 25ms was chosen, indicating muscle activity needed to remain above the initial 
threshold valued for at least 25ms. It is believed the double threshold method reduces the 
likelihood of identifying a false onset. The inclusion of the both amplitude and temporal aspects 
would seem to satisfy worries regarding Type I (false onset) or Type II (delay in onset) concerns 
(De Luca, 1997; Kamen & Gabriel, 2010). 
60 
 
 To determine the normalized peak sEMG activity of the six muscles during the gait cycle, 
muscle activity was trimmed to only include right stance phase (i.e. right foot strike to right toe-
off). Once identified, peak activity during the stance phase was normalized relative to the 
maximal voluntary isometric contraction (MVIC) for the respective muscle.  
Footwear 
 Four footwear conditions were utilized: (a) barefoot, (b) textured insole, (c) minimalist, 
zero-drop running shoe, and (d) textured insole inserted into the minimalist, zero-drop running 
shoe. During the barefoot condition, participants were completely barefoot with nothing to 
interfere with the foot-floor interface. The minimalist, zero-drop running shoe consisted of low-
top Adidas AdiPure Adapt (Adidas, Portland OR, USA) (Figure 13). This option was chosen due 
to the flat style design, which minimizes any differences in the vertical height of the heel and 
toes. Elevated heels have been previously shown to alter posture, gait kinematics, and muscle 
activity; however, at what point heel height differentials become significant is not yet understood 
(Aghazadeh & Lu, 1994; Bird et al., 2003; Gefen et al., 2002; C.-M. Lee et al., 2001; Murley et 
al., 2009). Therefore, as an elevated heel is common in most mass-production footwear, 
participants were not allowed to utilize their own footwear. The textured insole consisted of a 
model utilized by Hatton and colleagues (2011), which has previously been shown to result in 
significant reductions in postural sway. The textured insert was a 3 mm thick ethylene-vinyl 
acetate (EVA) insert cut to the foot size from a commercially available outsole material (Evalite 
Pyramids EVA, 3mm thickness, Algeos Ltd., Dubai, UAE). The material consisted of small 
pyramidal peaks with center-to-center distances of approximately 2.5mm (Figure 14). The insert 
was worn against the plantar surface of the foot. During insole only trials, a modified sandal was 
designed to minimize complications with fitting the insole to each participant while allowing for 
61 
 
natural foot motion (Figure 15). The combined effect resulted in minimal movement between the 
insole and the foot and reduced foot slippage during ground contact. During the combined insole 
and footwear condition, the insert was inserted directly into the shoe, negating the need for any 
additional material. All insoles were previously molded by the author to a matching shoe size. 
Finally, the footwear and textured inserts were only worn during data collection trials.  
 
 
Figure 13. Footwear Utilized in the Present Study. 
 
62 
 
 
Figure 14.  Textured insole utilized in the present study. 
 
 Figure 15. Huarache sandal design utilized in the present study for insole only conditions.  
 
Design & Procedures 
 Participants were asked to report to the Lower Extremity Laboratory for two days of 
testing, with the second session conducted within 72 hours of the first. Upon arrival on the first 
day, the participant was provided with an Auburn University Institutional Review Board 
approved informed consent form and health history questionnaire (Appendices A & B). By 
signing the informed consent, the participants indicated voluntary participation in the study. The 
health history questionnaire was used as a screening component to eliminate individuals who  
63 
 
have (1) any current or recent injury (<1 year) to the lower extremity, pelvic, low back, or trunk 
injury, (2) any previous injury that could jeopardize the successful performance of the requisite 
tasks, (3) any allergies to adhesives or adhesive type products, and/ or (4) any known balance 
deficits or inner ear disturbances. Upon meeting criteria for participation and indicated consent, 
participants were asked to try on the footwear to ensure proper shoe size selection. Next, the 
mass and height of the participant was measured for software inputs and normalization purposes. 
Additional anthropometric information (i.e. leg length, arch height index) was collected for 
future normalization purposes. While the aforementioned preparation steps were completed on 
the first visit, the following procedures were completed during both visits.  
All participants were asked to change into a compression outfit and discard any jewelry 
so as to minimize any potential noise for the motion capture system. The skin of the participant 
was then prepared for sEMG electrodes. Preparation was conducted in accordance with the 
recommendations of the Surface Electromyography for the Non-invasive Assessment of Muscles 
(SENIAM). The skin area around the electrode placement site was shaved, abraded, and cleaned 
with a 70% alcohol solution to reduce the likelihood of electrical impedance. Electrodes were 
placed 2-4 cm from the innervation zone, between the myotendinous junction, and parallel to the 
direction of the muscle fibers for the muscles of interest (De Luca, 1997; Kamen & Gabriel, 
2010; Nigg & Herzog, 2007). The lumbar portion of the right erector spinae, bilateral gluteus 
medius, right lateral hamstrings (e.g. biceps femoris), right soleus, and right peroneus longus 
were identified by palpation and manual muscle testing. Furthermore, anatomical reference 
charts were utilized to ensure proper placement. A ground electrode was placed on the body 
aspect of the tibial tuberosity. A thin, nylon type wrap was placed over the electrodes and limb 
segment to assist in preventing electrode movement. An MVIC was then performed and used for 
64 
 
normalization purposes (Kendall, 2005). An MVIC for the erector spinae was recorded during a 
prone lumbar extension task. An MVIC for the right and left gluteus medius was recorded during 
separate, but identical movements (side-laying, hip abduction). An MVIC for the biceps femoris 
was recorded during prone, knee flexion. The MVIC for the soleus and peroneus longus were 
recorded simultaneously during a seated, plantarflexion task with knees bent to 90?. 
 The final step of preparation was to affix 42 markers to the skin and clothing of the 
participant with double-sided tape. Marker placement was in accordance with the modified 
Vicon Plug-In-Gait model (Table 2). Next, a static capture was performed with the participants 
standing in the center of the capture volume in an anatomical neutral position, which is used as a 
reference point for the software. Marker locations and limb segments were then labeled in 
accordance with the modified Vicon Plug-In-Gait Model.  
Procedures 
 During the initial visit, ordering of conditions were randomized based on spatiotemporal 
variables (i.e. altered cadence, altered velocity). During the first day of data collection, only one 
variable was manipulated, while during the second day the other variable was manipulated. Next, 
footwear order was randomized (barefoot, shod, textured insole, shoe with textured insole). The 
initial trials for each condition were Normal walking, followed by a randomization of either 
125% or 75% of normal cadence or velocity. Randomization of conditions occurred through the 
use of a random number generator. After ordering of the conditions was performed, each 
participant was provided a minimum of three minutes in which to acclimate to the specific 
footwear condition. During this period, participants were asked to free-walk through the capture 
volume. After the acclimation period, but before each separate condition, participants were asked 
to practice the given condition to ensure compliance with velocity or cadence requirements. 
65 
 
Verbal feedback was provided by the author so the participant knew how to adjust to the given 
parameters. Each condition was repeated until three trials were accurately completed. Rest was 
provided pro re nata between trials or conditions. 
Experimental Design & Data Analysis 
Relative Phase Calculations  
 Relative phase calculations were determined based on methodology utilized in previous 
relative phase literature (Y. Huang et al., 2010; C. J. Lamoth, Beek, et al., 2002; Stergiou, 2004). 
Kinematic data was initially filtered using a dual-pass Butterworth filter. Next, phase portraits 
for pelvic, thoracic, and leg movements were projected onto a time series of displacement and 
velocity into a phase plane. The origin for each axis was reset to a midpoint between the 
extremes of each axis, with values normalized to the relative maximum and minimum values for 
displacement and velocity. Finally, the phase angle was calculated through a horizontal reference 
line, originating at the origin, and a projected ray intersecting the time point of interest. The 
angle between the reference line and ray defined the phase angle for a single segment in time. 
For the purposes of the present project, segmental movements were defined in relation to the 
right foot initial contact.  
The subtraction of segmental phase angles provided the continuous estimate of relative 
phase, as illustrated in Equation 1.  Relative phase values that are zero indicate the two segments 
are in-phase or synchronous, whereas relative phase values of 180? indicate asynchronous or out-
of-phase movements (Y. Huang et al., 2010; C. J. Lamoth, Beek, et al., 2002; Stergiou, 2004). 
Positive values indicate that the proximal segment is ahead of the distal segment, with negative 
values indicating the opposite (Stergiou, 2004). Furthermore, the slope of the relative phase 
provides information concerning segmental velocity, with a positive slope indicating the distal 
66 
 
segment is advancing faster in space and a negative slope indicating the proximal segment is 
advancing faster in space (Stergiou, 2004). Likewise, the local minimum and maximum provide 
information regarding changes in segmental coordination.  
                                                    
Equation 1: Relative Phase Calculation (Y. Huang et al., 2010; C. J. Lamoth, Meijer, et al., 
2002).  
Finally, as relative phase calculations are circular measures, it was necessary to take the 
absolute value of the relative phase measure and then calculate linear means and standard 
deviations (S.D.) using circular statistics, thereby avoiding phase wrapping and providing a 
single number for analysis (Y. Huang et al., 2010; C. J. Lamoth, Beek, et al., 2002; Stergiou, 
2004). The mean absolute relative phase (RP mean) was calculated as the summation of the mean 
curve by averaging the absolute values of the ensemble curve points (Equation 2) (Stergiou, 
2004). In Equation 2, N is the number of points in the relative phase mean ensemble and ? relative 
phase is the relative relationship between two segments (Equation 1). Finally, the deviation phase 
(RP SD) was another method of determining segmental coordination and the stability of the 
neuromuscular system (Equation 3) (Stergiou, 2004). In the calculation of RP SD, N is the 
number of points in the relative phase mean ensemble and SD is the standard deviation of the 
mean ensemble at the ith point (Stergiou, 2004). 
        ?|               | 
 
   
 
Equation 2: Mean Absolute Relative Phase (Stergiou, 2004).  
      ? |   |
 
   
  
Equation 3: Relative Phase Standard Deviation (Stergiou, 2004).  
67 
 
Statistical Analyses 
 All statistical analyses were conducted using SPSS software (version 18.0, SPSS Inc. 
Chicago, IL, USA) and an alpha level of statistical significance set apriori at ? ?0.05. All data 
was stored in Excel databases and imported into SPSS for analysis. To investigate the effects of 
footwear on gait kinematics during Normal walking, the kinematic data was analyzed using a 
two-way repeated measure MANOVA. The independent variables were footwear types, with 
four levels (barefoot, shod, textured insole, shod with textured insole), and gender with two 
levels. The dependent variables were stride length (SL), RP mean for the segments of interests 
(thorax-pelvis, thorax-leg, pelvis-leg), RP SD for the segments of interests (thorax-pelvis, thorax-
leg, pelvis-leg), and rotational amplitudes (thorax, pelvis, leg).  
 To investigate the effects of walking velocity on gait kinematics during shod walking, the 
kinematic data was analyzed with a two-way repeated measure MANOVA. The independent 
variables were walking velocity, with three levels (Normal, 125% of Normal velocity, and 75% 
of Normal velocity), and gender with two levels. The dependent variables were stride length 
(SL), cadence (SR), RP mean for the segments of interests (thorax-pelvis, thorax-leg, pelvis-leg), 
RP SD for the segments of interests (thorax-pelvis, thorax-leg, pelvis-leg), and rotational 
amplitudes (thorax, pelvis, leg). 
 To investigate the effects of cadence on gait kinematics during shod walking, the 
kinematic data was analyzed with a two-way repeated measure MANOVA. The independent 
variables were walking cadence, with three levels (Normal, 125% of Normal cadence, and 75% 
of Normal cadence), and gender with two levels. The dependent variables were stride length 
(SL), velocity, RP mean for the segments of interests (thorax-pelvis, thorax-leg, pelvis-leg), RP SD 
68 
 
for the segments of interests (thorax-pelvis, thorax-leg, pelvis-leg), and rotational amplitudes 
(thorax, pelvis, leg). 
 To investigate the effects of footwear on sEMG during Normal walking, the sEMG data 
was analyzed using a two-way repeated measure MANOVA. The independent variables were 
footwear types, with four levels (barefoot, shod, textured insole, shod with textured insole), and 
gender with two levels. The dependent variables were onset of muscular activity for the right 
erector spinae, right gluteus medius, left gluteus medius, right biceps femoris, right soleus, and 
right peroneus longus prior to foot strike, as well as the normalized peak sEMG of the 
aforementioned muscles during the stance phase of gait.  
To investigate the effects of walking velocity sEMG during shod walking, the sEMG data 
was analyzed with a two-way repeated measure MANOVA. The independent variables were 
walking velocity, with three levels (Normal, 125% of Normal velocity, and 75% of Normal 
velocity), and gender with two levels. The dependent variables were onset of muscular activity 
for the right erector spinae, right gluteus medius, left gluteus medius, right biceps femoris, right 
soleus, and right peroneus longus prior to foot strike, as well as the normalized peak sEMG of 
the aforementioned muscles during the stance phase of gait. 
To investigate the effects of cadence on sEMG during shod walking, the sEMG data was 
analyzed with a two-way repeated measure MANOVA. The independent variables were walking 
cadence, with three levels (Normal, 125% of Normal cadence, and 75% of Normal cadence), and 
gender with two levels. The dependent variables were onset of muscular activity for the right 
erector spinae, right gluteus medius, left gluteus medius, right biceps femoris, right soleus, and 
right peroneus longus prior to foot strike, as well as the normalized peak sEMG of the 
aforementioned muscles during the stance phase of gait. 
69 
 
For all statistical analyses, post hoc repeated measure ANOVA?s were performed for 
each dependent variable of the repeated measure MANOVA that demonstrated statistical 
significance. Follow-up pairwise comparisons were then performed for significant findings from 
the ANOVA analyses to determine statistical significance between the levels of the independent 
variables.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
70 
 
Chapter IV. Results 
 
 The purpose of this study was to better understand the effects of footwear on relative 
phase differences and muscle activation during walking. Specifically, this study investigated : (1) 
the effects of velocity and cadence on lumbopelvic rhythm during overground walking; (2) the 
effects of footwear on lumbopelvic rhythm during gait; (3) the effect of enhanced tactile 
feedback (i.e. barefoot, textured insole) on the onset of muscle activity at foot strike; (4) the 
effect of spatiotemporal and footwear variation on the magnitude of lumbar (i.e. erector spinae) 
and lower extremity muscular activity (e.g. gluteus medius, biceps femoris, soleus, peroneus 
longus) during the stance phase of gait, and (5) if kinematics or muscle activity patterns display 
significant gender differences. While human walking has been extensively studied in previous 
research, the relationship between trunk, pelvic, and leg movements are not yet fully understood 
(Crosbie & Vachalathiti, 1997; Crosbie et al., 1997a, 1997b; Y. Huang et al., 2010; Y. P. Huang 
et al., 2011; C. J. Lamoth, Beek, et al., 2002; C. J. Lamoth, Meijer, et al., 2002; C. J. C. Lamoth 
et al., 2006; Taylor et al., 2003; Taylor et al., 1999; van Emmerik & Wagenaar, 1996; W. Wu et 
al., 2002; W. Wu et al., 2004; W. H. Wu et al., 2008).  Likewise, while footwear has previously 
been shown to alter lower extremity joint kinematics and muscle activity, the extent to which 
tactile feedback is important in modulating walking is an area of emerging interest to researchers 
and clinicians. (Bird et al., 2003; Chen et al., 1995; Kelleher et al., 2010; Maki et al., 1999; 
Murley et al., 2009; Wakeling & Liphardt, 2006; Wakeling et al., 2003). The following chapter 
presents results of the current study in the following sections: (1) participant demographics; (2) 
the effect of walking velocity and gender on relative phase differences and muscle activation 
patterns; (3) the effect of cadence and gender on relative phase differences and muscle 
activation; and (4) the effect of footwear and gender on relative phase differences and muscle 
71 
 
activation patterns. To eliminate the potential of confounded results, data from the shod 
condition only was utilized to examine velocity and cadence effects. To analyze the footwear 
effects on kinematics and muscle activity, only the normal (e.g. self-selected) condition was 
included in the analyses.  
Section 1: Participant Demographics 
 One-hundred-twelve participants volunteered for participation in the current study, with 
fifty-nine participants meeting initial qualifications (Appendix A). Upon completion of the 
initial screening, each participant signed an Informed Consent document previously approved by 
the Auburn University Institutional Review Board for Human Subjects (Appendix B). Of the 
fifty-nine participants, four participants did not return for either testing session, one did not 
complete both testing sessions, and four were excluded due to erroneous data; therefore, fifty 
participants (25 men, 25 women) completed both testing sessions and were included in data 
analysis. Demographic information for the participants? is included in Table 3. 
 
 
 
 
 
 
 
 
 
.  
72 
 
Table 3 
Participant Demographics 
  
All Participants Mean Standard Deviation 
Age (years) 23.1 2.8 
Height (m) 1.72 0.11 
Mass (kg) 74.3 17.8 
Leg Length (cm) 86.2 5.7 
Females Mean Standard Deviation 
Age (years) 23.6 3.3 
Height (m) 1.64 0.06 
Mass (kg) 64.7 17.6 
Leg Length (cm) 82.8 4.4 
Males Mean Standard Deviation 
Age (years) 23.0 2.5 
Height (m) 1.81 0.07 
Mass (kg) 83.7 11.7 
Leg Length (cm) 89.4 4.8 
      
 
 
Section 2: Walking Velocity Effects 
 
 To understand the effects of walking velocity on relative phase differences during 
overground walking, velocity was manipulated on three levels (Normal, 125% of Normal, and 
75% of Normal) during one test session. Normal was self-selected by the participant. To 
determine differences between conditions, a 3 (condition) x 2 (gender) repeated measures 
MANOVA was performed for eleven dependent kinematic variables (Table 4). A significant 
(p<0.001) Box?s M test was found, therefore, the application of the more robust Pillai?s Trace 
multivariate test was utilized for interpretation of output statistics. 
 
 
 
 
 
 
 
73 
 
Table 4 ? Kinematic Variables: Velocity Effects 
Independent Variables Dependent Variables 
Gender Relative Phase mean 
- Male - Trunk-Pelvis (TP mean) 
- Female - Trunk-Right Leg (TL mean) 
Condition - Trunk-Left Leg (PL mean) 
- Normal Relative Phase SD 
- 125% of Normal - Trunk-Pelvis (TP SD) 
- 75% of Normal - Trunk-Right Leg (TL SD) 
 - Trunk-Left Leg (PL SD) 
Euler Angles (at foot strike) 
- Pelvis 
- Trunk 
- Right Leg 
Spatiotemporal Variables 
- Cadence 
- Normalized stride length 
 
 The multivariate repeated measures ANOVA with a p value set apriori at < 0.05 yielded 
a non-significant Gender*Condition interaction and a non-significant Gender effect; however, 
significant Condition (? Pillai = 0.998, F(22,27) = 656.053, p < 0.001, ?2= 0.998, Power = 1.000) 
differences were found. 
Spatiotemporal Variables: Condition Differences 
 Average walking velocity for all participants was 1.28 ms-1, 1.59 ms-1, and 0.96 ms-1 for 
the Normal, 125%, and 75% conditions, respectively. Follow-up ANOVAs found a significant 
difference for cadence (F(2,96) = 965.977, p < 0.001, ?2= 0.953, Power = 1.000) and normalized 
stride length (F(2,96) = 730.041, p < 0.001, ?2= 0.935, Power = 1.000). Post hoc analyses using 
the Least Significant Differences (LSD) found a significant difference (p<0.001) between 
cadence for all three conditions (Figure 16). Cadence was highest during the 125% velocity 
condition and lowest during the 75% velocity trials. Height normalized stride length was also 
found to be significantly different (p<0.001) between all conditions (Figure 17). Normalized 
74 
 
stride length was longest during the 125% velocity condition and lowest during the 75% velocity 
trials. 
  
Figure 16 ? Effect of walking velocity on cadence. Notations (*) indicate 
significance at (p<0.05). 
 
 
 
* 
* 
* 
75 
 
 
 
Figure 17 ? Effect of walking velocity on normalized stride length. Notations (*) 
indicate significance at (p<0.05). 
 
Spatiotemporal Variables: Gender Differences 
Because of the non-significant repeated measures MANOVA gender effect, follow-up 
ANOVAs were contraindicated, indicating gender had no effect on any of the spatiotemporal 
variables (Cadence, Stride length) when velocity was altered. 
Kinematics 
 Phase portraits were created as described in Chapter III and calculated using a custom 
MATLAB program (MATLAB R2013a Student Edition, Mathworks, Natick, MA, USA). Figure 
18 provides an example of a phase portrait for the pelvis during a full-stride (right toe-off to right 
toe-off) as calculated in the present study. The present study determined stride events based upon 
right foot events only (right foot strike, right toe-off) for all participants. From the phase portrait, 
phase angles were calculated for the trunk, pelvis, and leg at foot strike for each trial, after which 
the subtraction of the phase angles was performed according to Equation 1, with the difference 
* 
* 
* 
76 
 
equaling the relative phase value for the two segments. Relative phase measures range in value 
from 0? - 180?, with larger values implying greater asynchronicity, or out-of-phase relationships, 
between segments. Conversely, smaller values imply a more synchronous, or in-phase 
relationship between segments. Because relative phase is a circular measure, circular statistics 
were then applied (Equation 2 & 3) prior to the application of the linear statistics, resulting in 
three Relative Phase mean measurements (thorax-pelvis (TP mean ), thorax-leg (TL mean ), pelvis-leg 
(PL mean )) and three Relative Phase SD measurements for the segments of interests (thorax-pelvis 
(TP SD ), thorax-leg (TL SD ), pelvis-leg (PL SD )) in addition to the Euler Angles for the Pelvis, 
Trunk, and Right Leg at foot strike. The Relative Phase SD measure provides a measure of the 
amount of variability in the Relative Phase mean measure, or stated in another manner, the amount 
of variability in the synchronicity between segments.  
 Figure 18 ? Example Phase Portrait of the Pelvis during one-stride.  
 
 
 
77 
 
Kinematics: Condition Differences 
 Trunk-Pelvis mean 
Follow-up ANOVAs found a significant difference for TP mean (F(2,96) = 41.079, p < 
0.001, ?2= 0.461, Power = 1.000) with post hoc analyses using the LSD criterion indicating that 
the 75% velocity condition displayed a significantly greater (p<0.001) TP mean than the other 
conditions, while Normal was also significantly (p=0.008) greater than 125% velocity (Figure 
19). Results indicate that the pelvis and trunk were most out-of-phase, indicating greater 
asynchronicity, during the 75% velocity condition, relative to Normal and 125% velocity, and 
more out-of-phase during the Normal velocity than at 125% velocity. Follow-up ANOVA?s 
revealed no significant velocity effect on TP SD, indicating that the relationship between the trunk 
and pelvis do not become more variable as velocity increases from 75% of Normal to 125% of 
Normal.  
 
Figure 19 ? Effect of walking velocity on TP mean. Notations (*? ?) indicate 
significance at (p<0.05). 
 
? 
? 
* 
78 
 
Pelvis Rotation at Foot Strike 
Follow-up ANOVAs found a significant effect of velocity on the magnitude of pelvic 
rotation (F(2,96) = 354.773, p = 0.009, ?2= 0.094, Power = 0.800) with post-hoc analyses using a 
LSD criterion indicating that the 75% velocity condition was significantly lower than the Normal 
condition (p=0.004), yet no difference between Normal and 125% velocity or 125% velocity and 
75% velocity was noted. Results indicate that participants exhibited the greatest left transverse 
pelvic rotation at right foot strike during Normal walking while the pelvis was in a relatively 
neutral position at foot strike during 75% of Normal walking (Figure 20). 
 
Figure 20 ? Effect of walking velocity on pelvic rotation. Notations ( ?) indicate 
significance at (p<0.05). 
 
Trunk Rotation at Foot Strike 
Follow-up ANOVAs found a significant effect of cadence on the magnitude of trunk 
rotation (F(2,96) = 4.07, p = 0.020, ?2= 0.078, Power = 0.711) with post-hoc analyses using a 
? 
79 
 
LSD criterion indicating that the 75% velocity condition was significantly lower than 125% 
(p=0.004) and Normal condition (p=0.043), yet no difference between Normal and 125%, 
indicating the trunk was more rotated to the right at foot strike during Normal or 125% velocity 
walking than during 75% velocity (Figure 21).  
 
Figure 21 ? Effect of walking velocity on trunk rotation. Notations (? ?) indicate 
significance at (p<0.05). 
 
Non-significant Dependent Kinematic Variables During Varying Velocities 
Follow-up ANOVAs found no significant condition effect for the following kinematic 
variables (Table 5); indicating that changes in velocity did not significantly affect the 
participants for these dependent variables when velocity was altered at three levels (Normal, 
125%, 75%).  
 
 
? ? 
80 
 
Table 5 - Non-significant condition effects for kinematics when velocity was varied. 
  F df N p ?2 Power 
TP sd  2.247 2 96 0.111 0.045 0.448 
TL mean  2.391 2 96 0.097 0.047 0.472 
TL sd  1.289 2 96 0.280 0.026 0.273 
PL mean  2.695 2 96 0.073 0.053 0.522 
PL sd  2.825 2 96 0.064 0.056 0.543 
Leg Rotation 0.496 2 96 0.610 0.010 0.129 
              
 
Kinematics: Gender Differences 
Because of the non-significant repeated measures MANOVA gender effect, follow-up 
ANOVAs were contraindicated, indicating gender had no effect on any of the kinematic 
dependent variables (TP mean, TP sd, TL mean, TL sd, PL mean, PL sd, ? pelvis, ? trunk, ? leg) when 
velocity was altered. 
Surface Electromyography 
 
 To understand the effects of walking velocity on muscle activity during overground 
walking, electromyography (EMG) was recorded at 1500 HZ (Figure 12) using six pairs of 
bipolar Ag-AgCL surface electrodes (Trace-Rite, Bio-Detek Inc. Pawtucket, RI, USA), placed 
longitudinally over the bellies of six muscle (Table 7). To determine onset of muscle activity, a 
custom MATLAB program (MATLAB R2013a Student Edition, Mathworks, Natick, MA, USA) 
was written using the double-threshold technique as described in Chapter III, thereby 
determining the difference in time from onset of muscle activity and right foot strike. Visual 
inspections of the full-wave rectified signal, in conjunction with kinematic data, were also 
performed. Normalized peak EMG during the stance phase was also calculated using a custom 
MATLAB program, with all signals normalized to the maximal voluntary isometric contraction 
(MVIC) of the respective muscles. Though electrode placement sites were outlined during the 
81 
 
initial session, MVIC?s were performed at the beginning of each session for normalization 
purposes. To determine differences between conditions, a 3 (condition) x 2 (gender) repeated 
measures MANOVA was performed for twelve dependent variables (Table 6). A significant 
(p<0.001) Box?s M test was found, therefore, the application of the more robust Pillai?s Trace 
multivariate test was utilized for interpretation of output statistics. The multivariate repeated 
measures ANOVA with a p value set apriori at < 0.05 yielded a non-significant 
Gender*Condition interaction and a non-significant Gender effect; however, a significant 
Condition (? Pillai = 0.932, F(24,24) = 13.645, p < 0.001, ?2= 0.932, Power = 1.000) effect was 
found. 
Table 6 ? sEMG Variables 
Independent Variables Dependent Variables 
Gender Peak sEMG during Stance Phase 
- Male - Peroneus longus (PerL peak) 
- Female - Soleus (SOL peak) 
Condition - Biceps femoris (BF peak) 
- Normal - Right Gluteus medius (rGM peak) 
- 125% of Normal - Left Gluteus medius (lGM peak) 
- 75% of Normal - Erector spinae (ES peak) 
 Onset of Muscle Activity at Foot Strike 
- Peroneus longus (PerL onset) 
- Soleus (SOL onset) 
- Biceps femoris (BF onset) 
- Right Gluteus medius (rGM onset) 
- Left Gluteus medius (lGM onset) 
- Erector spinae (ES onset) 
 
 
Surface Electromyography: Condition Differences 
  
 Peak Peroneus longus 
 
Follow-up ANOVAs found a significant difference for PerL peak (F(2,94) = 69.815, p< 
0.001, ?2= 0.598, Power = 1.000) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the PerL peak during the 125% velocity condition was significantly 
82 
 
greater than Normal (p<0.001) and 75% (p<0.001). Furthermore, Normal was significantly 
greater than 75% velocity (p=0.013) indicating that alterations in velocity have a significant 
effect on peroneus longus activity during the stance phase of walking (Figure 22). 
 
 
 
Figure 22 ? Effect of walking velocity on PerL peak during stance. Notations (*) 
indicate significance at (p<0.05). 
 
 Peak Soleus 
 
Follow-up ANOVAs found a significant difference for SOL peak (F(2,94) = 20.79, p < 
0.001, ?2= 0.307, Power = 1.000) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the SOL peak was significantly greater during the 125% velocity 
condition than Normal (p<0.001) and 75% (p<0.001), though Normal and 75% velocity were 
not significantly different, indicating that velocities above normal have a significant effect on 
SOL peak (Figure 23).  
* 
* 
* 
83 
 
 
 
Figure 23 ? Effect of walking velocity on SOL peak during stance. Notations (? ?) 
indicate significance at (p<0.05). 
 
 Peak Biceps femoris 
 
Follow-up ANOVAs found a significant difference for BF peak (F(2,94) = 8.737, p = 
0.003, ?2= 0.157, Power = 0.966) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the BF peak during the stance phase was significantly different between 
all three velocity conditions. Specifically, the 125% velocity condition was significantly greater 
than Normal (p=0.021) and 75% (p=0.001) while Normal was significantly greater than 75% 
(p=0.002) indicating that alterations in velocity have a significant effect on biceps femoris 
activity during the stance phase of walking (Figure 24). 
? 
? 
84 
 
  
Figure 24 ? Effect of walking velocity on BF Peak during stance. Notations (*) 
indicate significance at (p<0.05). 
 
 Peak Right Gluteus medius 
 
Follow-up ANOVAs yielded a significant difference for rGM peak (F(2,94) = 16.222, p < 
0.001, ?2= 0.257, Power = 0.999) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the rGM peak during the stance phase was significantly different between 
all three velocity conditions. Specifically, the 125% velocity condition was significantly greater 
than Normal (p<0.001) and 75% (p=0.001) while Normal was significantly greater than 75% 
(p=0.027) indicating that alterations in velocity have a significant effect on right gluteus medius 
activity during the stance phase of walking (Figure 25). 
* 
* 
* 
85 
 
 
 
Figure 25 ? Effect of walking velocity on rGM Peak during stance. Notations (*) 
indicate significance at (p<0.05). 
 
 Peak Left Gluteus medius 
Follow-up ANOVAs yielded a significant difference for lGM peak (F(2,94) = 39.685, p < 
0.001, ?2= 0.458, Power = 1.000) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the lGM peak during the stance phase was significantly different between 
all three velocity conditions. Specifically, the 125% velocity condition was significantly greater 
than Normal (p<0.001) and 75% (p=0.001) while Normal was significantly greater than 75% 
(p=0.004) indicating that alterations in velocity have a significant effect on left gluteus medius 
activity during the stance phase of walking (Figure 26). 
* 
* 
* 
86 
 
 
 
Figure 26 ? Effect of walking velocity on lGM Peak during stance. Notations (*) 
indicate significance at (p<0.05). 
 
 Peak Erector spinae 
Follow-up ANOVAs yielded a significant difference for ES peak (F(2,94) = 11.478, p < 
0.001, ?2= 0.200, Power = 0.935) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the ES peak during the stance phase was significantly different between 
all three velocity conditions. Specifically, the 125% velocity condition was significantly greater 
than Normal (p=0.023) and 75% (p=0.001) while Normal was significantly greater than 75% 
(p<0.001) indicating that alterations in velocity have a significant effect on right erector spinae 
activity during the stance phase of walking (Figure 27). 
 
* 
* 
* 
87 
 
 
 
Figure 27 ? Effect of walking velocity on ES Peak during stance. Notations (*) 
indicate significance at (p<0.05). 
 
 
Non-significant Dependent sEMG Variables During Varying Velocities 
Follow-up ANOVAs found no significant condition effect for the following sEMG 
variables (Table 7); indicating that changes in velocity did not significantly affect participants for 
the dependent variables when velocity was altered at the three levels (Normal, 125%, 75%). 
 
Table 7 - Non-significant condition effects for sEMG when velocity was varied. 
  F df N p ?2 Power 
PerL onset  1.420 2 94 0.47 0.029 0.298 
Sol onset  1.533 2 94 0.221 0.032 0.319 
BF onset  0.860 2 94 0.405 0.018 0.176 
rGM onset  1.632 2 94 0.210 0.034 0.337 
lGM onset  2.892 2 94 0.060 0.058 0.553 
ES onset  2.989 2 94 0.055 0.060 0.568 
              
* 
* 
* 
88 
 
Surface Electromyography: Gender Differences 
Because of the non-significant repeated measures MANOVA gender effect, follow-up 
ANOVAs were contraindicated, indicating gender had no effect on any of the sEMG dependent 
variables (PerL peak, Sol peak, BF peak, rGM peak, lGM peak, ES peak, PerL onset, Sol onset, BF onset, rGM 
onset, lGM onset, ES onset) when velocity was altered. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
89 
 
Section 3: Cadence Effects 
 
 To understand the effects of walking cadence on relative phase differences during 
overground walking, cadence was manipulated on three levels (Normal, 125% of Normal, and 
75% of Normal) during a test session separate from the velocity session. To determine 
differences between conditions, a 3 (condition) x 2 (gender) repeated measures MANOVA was 
performed for eleven dependent kinematic variables (Table 8). A significant (p<0.001) Box?s M 
test was found, therefore, the application of the more robust Pillai?s Trace multivariate test was 
utilized for interpretation of output statistics. 
Table 8 ? Kinematic Variables: Cadence Effects 
Independent Variables Dependent Variables 
Gender Relative Phase mean 
- Male - Trunk-Pelvis (TP mean) 
- Female - Trunk-Right Leg (TL mean) 
Condition - Trunk-Left Leg (PL mean) 
- Normal Cadence Relative Phase SD 
- 125% of Normal - Trunk-Pelvis (TP SD) 
- 75% of Normal - Trunk-Right Leg (TL SD) 
 - Trunk-Left Leg (PL SD) 
Euler Angles (at foot strike) 
- Pelvis 
- Trunk 
- Right Leg 
Spatiotemporal Variables 
- Velocity 
- Normalized stride length 
 
 The multivariate repeated measures ANOVA with a p value set apriori at < 0.05 yielded 
a non-significant Gender*Condition interaction and a non-significant Gender effect; however, 
significant Condition (? Pillai = 0.998, F(22,27) = 25.670, p < 0.001, ?2= 0.954, Power = 1.000) 
differences were found. 
 
 
90 
 
Spatiotemporal Variables: Condition Differences 
 Average cadence for all participants was 139.6 steps*min-1(sm-1), 112.5 sm-1, and 84.7 
sm-1 for the 125%, Normal, and 75% conditions, respectively. Follow-up ANOVAs found a 
significant difference for velocity (F(2,96) = 199.71, p < 0.001, ?2= 0.745, Power = 1.000) and 
normalized stride length (F(2,96) = 66.854, p < 0.001, ?2= 0.494, Power = 1.000). Post hoc 
analyses using the LSD criterion found a significant difference (p<0.001) between velocity for 
all three conditions (Figure 28). Velocity was highest during the 125% cadence condition and 
lowest during the 75% cadence trials. Normalized stride length was also found to be significantly 
different (p<0.001) between all conditions (Figure 29), with the 125% cadence condition 
exhibiting the longest stride length and 75% cadence trials exhibiting the shortest stride lengths. 
  
Figure 28 ? Effect of cadence on velocities. Notations (*) indicate significance at 
(p<0.05). 
 
 
* 
* 
* 
91 
 
 
 
Figure 29 ? Effect of cadence on normalized stride length. Notations (*) indicate 
significance at (p<0.05). 
 
Spatiotemporal Variables: Gender Differences 
 Because of the non-significant repeated measures MANOVA gender effect, 
follow-up ANOVAs were contraindicated, indicating gender had no effect on any of the 
spatiotemporal variables (Velocity, Stride length) when Cadence was altered. 
Kinematics: Condition Differences 
 Trunk-Pelvis mean & Trunk-Pelvis SD 
Follow-up ANOVAs found a significant difference for TP mean (F(2,96) = 23.771, p < 
0.001, ?2= 0.331, Power = 1.000) and TP SD (F(2,96) = 3.818, p = 0.025, ?2= 0.074, Power = 
0.681). Post hoc analyses using the LSD criterion found that the 75% cadence condition 
displayed a significantly greater (p<0.001) TP mean than 125% and Normal (Figure 30). Post hoc 
analyses also revealed that TP SD was significantly lower for 125% cadence than 75% cadence 
* 
* 
* 
92 
 
(p=0.006) only (Figures 31). Results indicate that the pelvis and trunk were most out-of-phase 
during the 75% cadence condition, relative to Normal and 125% cadence. 
 
 
 
Figure 30 ? Effect of walking cadence on TP mean. Notations (*?) indicate 
significance at (p<0.05). 
 
? 
* 
93 
 
 
 
Figure 31 ? Effect of walking cadence on TP SD. Notations (*) indicate 
significance at (p<0.05). 
  
 
 
Trunk-Leg mean & Trunk-Leg SD 
A follow-up ANOVA found a significant effect of cadence on TL mean (F(2,96) = 4.678, p 
= 0.012, ?2= 0.089, Power = 0.774) with post-hoc analyses using the LSD criterion indicating 
that the normal cadence condition was significantly lower than 125% cadence (p=0.002) and 
75% cadence (p=0.037), indicating the trunk and leg were significantly more synchronized 
during the Normal condition relative to the 125% cadence and 75% cadence  (Figure 32). A 
follow-up ANOVA revealed no significant difference for TL SD between any of the conditions 
(F(2,96) = 0.186, p = 0.931, ?2= 0.004, Power = 0.078), indicating that the variability of the TL 
were not significantly different across the cadence conditions. 
* 
94 
 
  
Figure 32 ? Effect of walking cadence on TL mean. Notations (*?) indicate 
significance at (p<0.05). 
 
 
 
 
 
 Pelvis-Leg mean & Pelvis-Leg SD 
A follow-up ANOVA found a significant effect of cadence on PL mean (F(2,96) = 11.963, 
p < 0.001, ?2= 0.200, Power = 0.994) with post-hoc analyses using a LSD criterion indicating 
that the Normal cadence condition was significant lower than 125% cadence (p=0.013) and 75% 
cadence (p<0.001) , while 125% cadence was significantly lower (p=0.027) than 75% cadence 
(Figure 33). A follow-up ANOVA revealed no significant difference for TL SD (F(2,96) = 2.747, 
p = 0.069, ?2= 0.054, Power = 0.531). Results indicate that the pelvis and leg were significantly 
more synchronized during the Normal condition relative to the 125% cadence and 75% cadence 
while variability in TP measures were not significantly altered by the cadence.  
* ? 
95 
 
 
 
Figure 33 ? Effect of walking cadence on PL mean. Notations (*? ?) indicate 
significance at (p<0.05). 
 
Pelvis Rotation at Foot Strike 
Follow-up ANOVAs found a significant effect of cadence on the magnitude of pelvic 
rotation (F(2,96) = 24.150, p < 0.001, ?2= 0.335, Power = 1.000) with post-hoc analyses using a 
LSD criterion indicating that the 125% cadence condition was significant greater than 75% 
cadence  and Normal condition (p<0.001), yet no difference between Normal or 75% cadence 
was noted. Results indicate that while the pelvis was exhibiting left transverse pelvic girdle 
rotation at foot strike when walking at Normal or 75% cadence, the pelvis was right transverse 
rotated at foot strike during the 125% cadence (Figure 34). 
* 
? 
? 
96 
 
 
Figure 34 ? Effect of walking cadence on pelvic rotation. Notations (*?) indicate 
significance at (p<0.05). 
 
Trunk Rotation at Foot Strike 
Follow-up ANOVAs found a significant effect of cadence on the magnitude of trunk 
rotation (F(2,96) = 67.904, p < 0.001, ?2= 0.583, Power = 1.000) with post-hoc analyses using a 
LSD criterion indicating that the 125% cadence condition was significant greater than 75%  
cadence and Normal condition (p<0.001), yet no difference between Normal and 75% cadence, 
indicating greater trunk rotation at foot strike when walking at the 125% cadence (Figure 35). 
Furthermore, while the trunk was rotated to the right at foot strike during the Normal and 75% 
cadence conditions, results indicate that the trunk was rotated to the left at the 125% cadence 
condition.  
 
? 
* 
97 
 
 
Figure 35 ? Effect of walking cadence on trunk rotation. Notations (*?) indicate 
significance at (p<0.05). 
  
Leg Rotation at Foot Strike 
Follow-up ANOVAs found a significant effect of cadence on the magnitude of pelvic 
rotation (F(2,96) = 112.777, p < 0.001, ?2= 0.701, Power = 1.000) with post-hoc analyses using a 
LSD criterion indicating that the 125% cadence condition was significantly greater than than 
75% cadence and Normal conditions (p<0.001), yet no difference between Normal or 75% 
cadence was noted, signifying greater external rotation at foot strike when walking at 125%  of 
Normal cadence (Figure 36). 
? 
* 
98 
 
 
Figure 36 ? Effect of walking cadence on leg rotation. Notations (*) indicate 
significance at (p<0.05). 
 
Kinematics: Gender Differences 
Because of the non-significant repeated measures MANOVA gender effect, follow-up 
ANOVAs were contraindicated, indicating gender had no effect on any of the kinematic 
dependent variables (TP mean, TP sd, TL mean, TL sd, PL mean, PL sd, ? pelvis, ? trunk, ? leg) when 
cadence was altered. 
Surface Electromyography 
  
 To determine muscle activity during overground walking, electromyography (EMG) was 
recorded at 1500 HZ (Figure 5) using six pairs of bipolar Ag-AgCL surface electrodes (Trace-
Rite, Bio-Detek Inc. Pawtucket, RI, USA), placed longitudinally over the bellies of six muscles 
(Table 7). A 3 (condition) x 2 (gender) repeated measures MANOVA was performed for twelve 
dependent variables (Table 7).To understand the effects of cadence on muscle activity during 
* 
* 
99 
 
overground walking, cadence was manipulated on three levels (Normal, 125% of Normal, and 
75% of Normal) during a single test session. A significant (p<0.001) Box?s M test was found, 
therefore, the application of the more robust Pillai?s Trace multivariate test was utilized for 
interpretation of output statistics. The multivariate repeated measures ANOVA with a p value set 
apriori at < 0.05 yielded a non-significant Gender*Condition interaction and a non-significant 
Gender effect; however, a significant Condition (? Pillai = 0.847, F(24,24) = 5.528, p < 0.001, 
?2= 0.947, Power = 1.000) effect was found. 
Surface Electromyography: Condition Differences 
  
 Peak Peroneus longus 
 
Follow-up ANOVAs found a significant difference for PerL peak (F(2,94) = 17.960, p < 
0.001, ?2= 0.276, Power = 1.000). Post hoc analyses using the LSD criterion determined that the 
PerL peak during the 125% cadence condition was significantly greater (p<0.001) than the 75% 
cadence and Normal Cadence (p<0.001) conditions (Figure 37).  
 
 
100 
 
 
Figure 37 ? Effect of cadence on PerL peak. Notations (*?) indicate significance at 
(p<0.05). 
 
 Peak Soleus 
 
Follow-up ANOVAs found a significant difference for SOL peak (F(2,94) = 50.475, p < 
0.001, ?2= 0.518, Power = 1.000). Post hoc analyses using the LSD criterion determined that the 
Sol peak during the 125% cadence condition was significantly greater than Normal (p<0.001) and 
75% cadence (p<0.001). Furthermore, Normal was significantly greater than 75% cadence 
(p<0.001) indicating that alterations in cadence have a significant effect on soleus activity during 
the stance phase of walking (Figure 38).  
* 
? 
101 
 
 
Figure 38 ? Effect of walking cadence on SOL Peak. Notations (* ?) indicate 
significance at (p<0.05). 
  
 
Peak Biceps femoris 
 
Follow-up ANOVAs found a significant difference for BF peak (F(2,94) = 22.052, p < 
0.001, ?2= 0.319, Power = 1.000) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the BF peak during the stance phase was significantly greater during 
125% cadence conditions than either Normal (p<0.001) or 75% cadence (p<0.001); however, 
Normal was not significantly different than 75% cadence indicating that alterations in cadence 
above Normal have a significant effect on biceps femoris activity during the stance phase of 
walking (Figure 39). 
 
 
 
* 
* 
* 
102 
 
 
 
Figure 39 ? Effect of walking cadence on BF Peak. Notations (* ?) indicate 
significance at (p<0.05). 
 
Peak Right Gluteus medius 
 
Follow-up ANOVAs yielded a significant difference for rGM peak (F(2,94) = 24.678, p < 
0.001, ?2= 0.344, Power = 1.00). Post hoc analyses using the LSD criterion determined that the 
rGM peak during the 125% cadence condition was significantly greater than during the Normal 
(p<0.001) and 75% cadence conditions (p<0.001) only, indicating that changes in cadence at 
lower levels did not have a significant effect on rGM peak, but changes in cadence above Normal 
did have a significant impact on peak muscle activity (Figure 40). 
* 
? 
103 
 
 
 
Figure 40 ? Effect of cadence on rGM Peak. Notations (*?) indicate significance at 
(p<0.05). 
 
 Peak Left Gluteus medius 
Follow-up ANOVAs yielded a significant difference for lGM peak (F(2,94) = 38.375, p < 
0.001, ?2= 0.449, Power = 1.000). Post hoc analyses using the LSD criterion determined that the 
lGM peak during the 125% cadence condition was significantly greater than the Normal (p<.001) 
and 75% cadence conditions (p<0.001) only, indicating that changes in cadence at lower levels 
did not have a significant effect on lGM peak, but changes in cadence above Normal did have a 
significant impact on peak muscle activity (Figure 41). 
* 
? 
 
104 
 
 
Figure 41 ? Effect of cadence on lGM Peak. Notations (*?) indicate significance at 
(p<0.05). 
  
Peak Erector spinae 
Follow-up ANOVAs yielded a significant difference for ES peak (F(2,94) = 6.960, p = 
0.002, ?2= 0.129, Power = 0.918) during the stance phase. Post hoc analyses using the LSD 
criterion determined that the ES peak during the stance phase was significantly greater during the 
125% cadence condition than Normal (p=0.030) and 75% cadence (p=0.001) only, indicating 
that alterations in cadence above Normal have a significant effect on right erector spinae activity 
during the stance phase of walking (Figure 42). 
? 
* 
105 
 
 
Figure 42 ? Effect of cadence on ES Peak. Notations (*?) indicate significance at 
(p<0.05). 
  
Non-significant Dependent sEMG Variables During Varying Cadences 
Follow-up ANOVAs found no significant condition effect for the following sEMG 
variables (Table 9); indicating that changes in cadence did not significantly affect participants for 
the dependent variables when cadence was altered at the three levels (Normal, 125%, 75%). 
Table 9 - Non-significant condition effects for sEMG when cadence was varied. 
  F df N p ?2 Power 
PerL onset  0.027 2 94 0.974 0.001 0.054 
Sol onset  1.144 2 94 0.323 0.024 0.246 
BF onset  0.597 2 94 0.561 0.013 0.146 
rGM onset  1.082 2 94 0.343 0.022 0.234 
lGM onset  2.010 2 94 0.140 0.041 0.406 
ES onset  2.183 2 94 0.118 0.044 0.436 
              
 
 
* 
? 
106 
 
Surface Electromyography: Gender Differences 
Because of the non-significant repeated measures MANOVA gender effect, follow-up 
ANOVAs were contraindicated, indicating gender had no effect on any of the sEMG dependent 
variables (PerL peak, Sol peak, BF peak, rGM peak, lGM peak, ES peak, PerL onset, Sol onset, BF onset, rGM 
onset, lGM onset, ES onset) when cadence was altered. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
107 
 
Section 4: Footwear Effects 
 
 To understand the effects of footwear on relative phase differences during overground 
walking, footwear was manipulated on four levels (Table 13) during two separate testing 
sessions. Only kinematics and sEMG data from the Normal conditions (velocity and cadence) 
were used to examine potential footwear effects during Normal walking (i.e. 125% and 75% 
conditions were excluded from analyses). To determine differences between conditions, a 4 
(condition) x 2 (gender) repeated measures MANOVA was performed for twelve dependent 
kinematic variables (Table 10). A significant (p<0.001) Box?s M test was found, therefore, the 
application of the more robust Pillai?s Trace multivariate test was utilized for interpretation of 
output statistics. 
Table 10 ? Kinematic Variables: Footwear Effects 
Independent Variables Dependent Variables 
Gender Relative Phase mean 
- Male - Trunk-Pelvis (TP mean) 
- Female - Trunk-Right Leg (TL mean) 
Condition - Trunk-Left Leg (PL mean) 
- Shod (SH) Relative Phase SD 
- Barefoot (BaFt) - Trunk-Pelvis (TP SD) 
- Insole only (IN) - Trunk-Right Leg (TL SD) 
- Insole with shoe (INSH) - Trunk-Left Leg (PL SD) 
 Euler Angles (at foot strike) 
- Pelvis 
- Trunk 
- Right Leg 
Spatiotemporal Variables 
- Velocity 
 - Cadence 
 - Normalized stride length 
  
The multivariate repeated measures ANOVA with a p value set apriori at < 0.05 yielded 
a non-significant Gender*Condition interaction; however, significant Gender (? Pillai = 0.767, 
108 
 
F(12,37) = 10.152, p < 0.001, ?2= 0.767, Power = 1.000) and Condition (? Pillai = 0.956, 
F(36,13) = 7.839, p < 0.001, ?2= 0.077, Power = 0.956) differences were found. 
Spatiotemporal Variables: Condition Differences 
 Follow-up ANOVAs found a significant footwear effect on normalized stride length 
(F(3,144) = 58.310, p < 0.001, ?2= 0.548, Power = 1.000), velocity (F(3,144) = 9.479, p < 0.001, 
?2= 0.164, Power = 0.997), and cadence (F(3,144) = 7.427, p < 0.001, ?2= 0.134, Power = 
0.984). Post hoc analyses using the LSD criterion found a significant difference (p<0.001) in 
normalized stride length between shod (SH) vs. barefoot (BaFt), SH vs. insole (IN), insole-shod 
(INSH) vs. BaFt, and INSH vs. IN. Differences in normalized stride length between SH vs. 
INSH and BaFt vs. IN were non-significant (Figure 43). 
  
 
Figure 43 ? Effect of footwear on normalized stride length during normal 
walking. Notations (*? ? ?) indicate significance at (p<0.05). 
 
 
Post hoc analyses using the LSD criterion found a significant difference (p?0.001) in 
normalized stride length between SH vs. BaF, SH vs. IN, INSH vs. BaFt, and INSH vs. IN. 
Differences in velocity between SH vs. INSH and BaFt vs. IN were non-significant (Figure 44). 
? * 
? ? 
109 
 
 
Figure 44 ? Effect of footwear on velocity during normal walking. Notations (*? 
? ?) indicate significance at (p<0.05). 
 
Post hoc analyses using the LSD criterion found a significant difference (p<0.001) in 
normalized stride length between SH vs. BaFt, BaFt vs. IN, and INSH vs. BaFt. Differences in 
cadence between SH vs. INSH, , INSH vs. IN, and SH vs. IN were non-significant (Figure 45). 
  
Figure 45 ? Effect of footwear on cadence during normal walking. Notations (*? 
? ) indicate significance at (p<0.05). 
 
? ? 
? * 
? * 
? 
110 
 
Spatiotemporal Variables: Gender Differences 
 Follow-up ANOVAs found a significant gender effect on normalized stride length 
(F(1,48) = 8.224, p= 0.006, ?2= 0.115, Power = 0.099) and cadence (F(1,48) = 17.683, p < 
0.001, ?2= 0.269, Power = 0.206); however, a non-significant difference in velocity (F(1,48) = 
0.737, p =0.395, ?2= 0.0015, Power = 0.086) was noted. Descriptive statistics revealed that 
females took significantly longer height normalized strides (Figure 46) and had a significantly 
higher cadence during Normal walking than males (Figure 47).  
 
  
Figure 46? Effect of gender on normalized stride length during normal walking 
when footwear is altered. Notations (*) indicate significance at (p<0.05). 
 
* 
* 
111 
 
  
Figure 47 ? Effect of gender on cadence during normal walking when footwear is 
altered. Notations (*) indicate significance at (p<0.05). 
 
Kinematics: Condition Differences  
 Pelvis-Leg mean  
A follow-up ANOVA found a significant footwear effect on PL mean (F(3,144) = 4.629, p 
= 0.004, ?2= 0.088, Power = 0.884) with post-hoc analyses using a LSD criterion yielding that 
the BaFt condition was significantly greater than SH (p=0.048)  and INSH (p = 0.008), but not 
IN, while IN was significantly greater than SH (p=0.022) and INSH (p=0.004) (Figure 48), 
indicating that the pelvis and leg were more asynchronous during the BF and IN conditions 
relative to the other conditions. 
* * 
112 
 
 
Figure 48 ? Effect of footwear on PL mean. Notations (*?  ? ?) indicate 
significance at (p<0.05). 
 
Non-significant Dependent Kinematic Variables During Different Footwear 
Follow-up ANOVAs found no significant condition effect for the following kinematic 
variables (Table 11); indicating footwear did not significantly affect these dependent variables 
when altered at the four levels (SH, BaFt, IN, INSH).  
Table 11 - Non-significant condition effects for kinematics when footwear was varied. 
  F df N p ?2 Power 
TP mean  0.289 3 144 0.833 0.006 0.105 
TP sd 0.961 3 144 0.413 0.020 0.258 
TL mean  2.583 3 144 0.056 0.051 0.625 
TL sd  2.274 3 144 0.082 0.045 0.565 
PL sd  0.536 3 144 0.645 0.011 0.153 
Pelvic Rotation 0.937 3 144 0.424 0.019 0.253 
Trunk Rotation 0.975 3 144 0.406 0.020 0.262 
Leg Rotation 0.148 3 144 0.931 00.003 0.077 
              
 
? * ?  ? 
113 
 
Kinematics: Gender Differences 
Trunk-Pelvis mean & Trunk-Pelvis SD 
 
Follow-up ANOVAs revealed a non-significant gender effect for TP mean (F(1,48) = 
3.439, p = 0.070, ?2= 0.067, Power = 0.094); however, a significant gender difference was noted 
for TP SD (F(1,48) = 6.973, p = 0.011, ?2= 0.127, Power = 0.481) , indicating males and females 
were significantly different in the level of variability between trunk and pelvis during 
overground walking when footwear was altered, with males exhibiting greater variability than 
females (Figures 49). 
 
Figure 49 ? Effect of gender on TP SD during different footwear. Notations (*) 
indicate significance at (p<0.05). 
 
 
 
 
* 
* 
114 
 
Non-significant Dependent Kinematic Variables between Males and Females During 
Altered Footwear 
Follow-up ANOVAs found no significant gender effect for the following kinematic 
variables (Table 12); indicating changes in footwear (SH, BaFt, IN, INSH) did not affect males 
or females to a greater extent for the following dependent variables. 
Table 12 ? Non-significant gender effects when footwear was varied. 
  F df N p ?2 Power 
TL mean  0.872 1 48 0.355 0.018 0.181 
TL sd  0.548 1 48 0.463 0.011 0.069 
PL mean  2.579 1 48 0.115 0.051 0.053 
PL sd  2.731 1 48 0.105 0.054 0.062 
Pelvic Rotation 0.195 1 48 0.661 0.004 0.086 
Trunk Rotation 1.838 1 48 0.181 0.037 0.170 
Leg Rotation 3.144 1 48 0.083 0.061 0.107 
              
 
Surface Electromyography 
 
 To understand the effects of footwear on muscle activity during overground walking, 
footwear was manipulated on four levels (Table 13) during two separate testing sessions. To 
determine differences between conditions, a 4 (condition) x 2 (gender) repeated measures 
MANOVA was performed for twelve dependent variables (Table 13). A significant (p<0.001) 
Box?s M test was found, therefore, the application of the more robust Pillai?s Trace multivariate 
test was utilized for interpretation of output statistics. The multivariate repeated measures 
ANOVA with a p value set apriori at < 0.05 yielded a non-significant Gender*Condition 
interaction. Furthermore, while there was a non-significant Gender effect (? Pillai = 0.369, 
F(12,36) = 1.757, p = 0.095, ?2= 0.369, Power = 0.764) and Condition effect (? Pillai = 0.835, 
F(36,12) = 1.689, p = 0.166, ?2= 0.835, Power = 0.625). 
 
115 
 
 
Table 13 ? sEMG Variables: Footwear Effects 
Independent Variables Dependent Variables 
Gender Peak sEMG during Stance Phase 
- Male - Peroneus longus (PerL peak) 
- Female - Soleus (SOL peak) 
Condition - Biceps femoris (BF peak) 
- Shod (SH) - Right Gluteus medius (rGM peak) 
- Barefoot (BaFt) - Left Gluteus medius (lGM peak) 
- Insole only (IN) - Erector spinae (ES peak) 
- Insole with shoe (INSH) Onset of Muscle Activity at Foot Strike 
 - Peroneus longus (PerL onset) 
- Soleus (SOL onset) 
- Biceps femoris (BF onset) 
- Right Gluteus medius (rGM onset) 
- Left Gluteus medius (lGM onset) 
- Erector spinae (ES onset) 
 
 
Surface Electromyography: Condition Differences 
   
Because of the non-significant repeated measures MANOVA condition effect, follow-up 
ANOVAs were contraindicated, indicating the different footwear conditions had no effect on any 
of the sEMG dependent variables (PerL peak, Sol peak, BF peak, rGM peak, lGM peak, ES peak, PerL 
onset, Sol onset, BF onset, rGM onset, lGM onset, ES onset). 
Surface Electromyography: Gender Differences 
 
Because of the non-significant repeated measures MANOVA gender effect, follow-up 
ANOVAs were contraindicated, indicating gender had no effect on any of the sEMG dependent 
variables (PerL peak, Sol peak, BF peak, rGM peak, lGM peak, ES peak, PerL onset, Sol onset, BF onset, rGM 
onset, lGM onset, ES onset) when cadence was altered. 
 
 
 
116 
 
Chapter V. Discussion 
 The purpose of this study was to better understand the effects of altered spatiotemporal 
variables and footwear on relative phase differences and muscle activation during walking. 
Specifically, this study investigated : (1) the effects of velocity and cadence on lumbopelvic 
rhythm during overground walking; (2) the effects of footwear on lumbopelvic rhythm during 
gait; (3) the effect of enhanced tactile feedback (i.e. barefoot, textured insole) on the onset of 
muscle activity at foot strike; (4) the effect of spatiotemporal and footwear variation on the 
magnitude of lumbar (i.e. erector spinae) and lower extremity muscular activity (e.g. gluteus 
medius, biceps femoris, soleus, peroneus longus) during the stance phase of gait, and (5) 
differences in the kinematics and electromyography between males and females. The following 
chapter has been divided into five sections, with the initial three sections addressing the overall 
research questions. Section 1 discusses the effect of walking velocity on relative phase 
differences and muscle activation patterns. Section 2 discusses the effect of cadence on relative 
phase differences and muscle activation. Section 3 discusses the effect of footwear and gender on 
relative phase differences and muscle activation patterns. Section 4 provides a summary of the 
combined findings in light of the five research questions. Section 5 discusses final conclusions 
from the present research and recommendations for future research.   
Section 1: Walking Velocity Effects 
 The purpose of this section of the study was to investigate the effect of altered velocity on 
spatiotemporal variables, relative phase differences, and electromyography at three velocities 
(Normal, 125% of Normal, 75% of Normal velocity). The kinematic variables of interest were 
height normalized stride length (SL), cadence, relative phase mean differences for the trunk-
pelvis, trunk-leg, and pelvis leg (TP mean, TL mean, PL mean, respectively), relative phase standard 
117 
 
deviation differences for the trunk-pelvis, trunk-leg, and pelvis-leg (TP sd, TL sd, PL sd, 
respectively), and angular positions of the pelvis, trunk, and leg at foot strike. Table 14 illustrates 
the spatiotemporal variables and relative phase differences for the altered velocity conditions as 
compared to the normal velocity condition.  
Table 14 - Effect of Velocity on all Kinematic Variables as 
Compared to Normal 
Variable 125% 75% 
SL ? ? 
Cadence ? ? 
TP mean ? ? 
TP sd ? ? 
TL mean ? ? 
TL sd ? ? 
PL mean ? ? 
PL sd ? ? 
? pelvis ? ? 
? thorax ? ? 
? leg ? ? 
Note.  ? = no significant difference, ? = increase, ? = decrease 
 
All indicated differences were significantly different (p <0.05) 
 
The sEMG variables of interest were the normalized peak activity during the stance phase 
of the peroneus longus (PerL peak), soleus (Sol peak), biceps femoris (BF peak), right gluteus medius 
(rGM peak), left gluteus medius (lGM peak), and erector spinae (ES peak). Furthermore, the time to 
onset muscle activity for the same six muscles at foot strike (PerL onset, Sol onset, BF onset, rGM 
onset, lGM onset, and ES onset) were also investigated. Table 15 illustrates the sEMG (normalized 
peak and time to onset for the respective six muscles) differences for the altered velocity 
conditions as compared to the normal velocity condition. 
 
118 
 
Table 15 - Effect of Velocity on all sEMG Variables as Compared 
to Normal 
Variable 125% 75% 
PerL peak  ? ? 
Sol peak  ? ? 
BF peak  ? ? 
rGM peak  ? ? 
lGM peak  ? ? 
ES peak  ? ? 
PerL onset  ? ? 
Sol onset  ? ? 
BF onset  ? ? 
rGM onset  ? ? 
lGM onset  ? ? 
ES onset  ? ? 
Note.  ? = no significant difference, ? = increase, ? = decrease 
 
All indicated differences were significantly different (p <0.05) 
 
Condition Differences in Spatiotemporal & Kinematic Variables During Altered Velocity 
 The present study found a significant difference in cadence between all velocity 
conditions (Figure 16), indicating that as velocity was increased from 75% of Normal, to 
Normal, to 125% of Normal, participants adjusted cadence by increasing the number of steps per 
minute. The present study also found a significant difference in height normalized stride length 
between all velocity conditions (Figure 17), indicating that as velocity was increased from 75% 
of Normal, to Normal, to 125% of Normal, participants adjusted stride length by taking longer 
height normalized strides. It was anticipated that as the requested velocity increased, the 
participants would assume either a cadence or stride length dominant strategy to achieve the 
imposed speed; however, it appears that participants adjusted both variables, to varying degrees, 
to achieve the chosen requested velocity.  To better understand the results of the present study, 
119 
 
an investigation of whether participants achieved the increased normalized stride length by 
increased hip flexion/knee extension or pelvic girdle rotation was considered. As indicated in 
Figure?s 21 & 22, there was a significant difference the amount of pelvic rotation and trunk 
rotation at foot contact, suggesting that the increase in normalized stride length may have 
resulted from a greater pelvic rotation during terminal swing. Conversely, as pelvic rotation was 
greatest during the Normal condition, normalized stride length may have been achieved by a 
longer excursion of the center of mass prior to foot contact.  Furthermore, the increased 
excursion must have occurred through an exaggerated toe-off which assisted in moving the foot 
to move forward more rapidly (i.e. increased cadence), yet the body displayed similar alignments 
at foot contact.  This supposition is further supported by the sEMG findings as the soleus 
produced, relative to the Normal condition, a diminished excitation during the 75% of Normal 
condition and an increased excitation during the 125% of Normal.  Furthermore, the significant 
increase in peak peroneus longus activity would also support this supposition, as the participant 
would presumably attempt to rapidly evert the foot to better load the head of the metatarsals (e.g. 
ball of the foot) and the hallux (e.g. big toe) during toe-off.   As velocity (and cadence) 
increased, the rate of eversion would occur faster due to shorter absolute ground contact times.  
Support of this theory also comes from the sEMG of the BF peak, which indicated a significant 
increase in peak activity as velocity was increased. Conversely, the increase in biceps femoris 
activity may have resulted from the longer normalized stride length as velocity increased. Of 
course, in order to draw a definitive conclusion, joint angles at toe off and kinetic data would 
need to be investigated. 
 In the present study, TP mean decreased as the velocity increased from 75% to 125%, 
indicating a more in-phase relationship between the trunk and pelvis during higher velocities 
120 
 
(Figure 20). These findings are opposite of previous research examining the effect of velocity on 
trunk-pelvis coordination during walking (Bruijn et al., 2008; Y. Huang et al., 2010; C. J. 
Lamoth, Beek, et al., 2002; Wagenaar & van Emmerik, 2000). Several factors may account for 
the difference in findings between previous research and the present study. First, the velocities 
utilized in the present study were much larger than those employed in previous studies (Bruijn et 
al., 2008; Y. Huang et al., 2010; C. J. Lamoth, Beek, et al., 2002).  Second, to the knowledge of 
the author, the present study is the first to examine the relative phase differences during 
overground walking with all previous research completed on treadmill walking. Though debated 
in the literature, several studies have noted significant biomechanical differences, particularly at 
the knee and hip, in treadmill walking relative to overground walking (Alton, Baldey, Caplan, & 
Morrissey, 1998; Dingwell, Cusumano, Cavanagh, & Sternad, 2001; S. J. Lee & Hidler, 2008; 
Parvataneni, Ploeg, Olney, & Brouwer, 2009). Furthermore, Dingwell et al. (2001) demonstrated 
significant reductions in both variability and local dynamic stability in treadmill walking relative 
to overground walking. Interestingly, in a study examining pelvic movements during overground 
walking, Crosbie and Vachalathiti (1997) reported that walking velocities above normal tended 
to produce a tight phase locking of movement patterns.   So, though the findings of the present 
study do not coincide with the findings of previous outcomes during treadmill walking, the 
present study did support the findings of other work performed on overground walking. 
The present study also found a significant increase in bilateral gluteus medius (Figure?s 
25 & 26) and erector spinae (Figure 27) activity during stance phase as velocity increased from 
75% of Normal conditions relative to other conditions. This finding is interesting considering the 
more synchronous relationship between the trunk and pelvis that was found in the present study; 
however, the increased activity of the lumbopelvichip (LPH) musculature may have arisen from 
121 
 
two separate areas. As stride length and velocity increased, step width (not reported) would have 
decreased, placing additional demands on the respective GM to assist in maintaining frontal 
plane pelvic neutrality during single-limb support (Perry, 1992; Whittle, 2003). Increased LPH 
activity may have also resulted from the phase-locking and increased co-contraction of 
lumbopelvichip musculature, which has been previously hypothesized to occur during 
overground walking at greater than normal velocities (Crosbie & Vachalathiti, 1997). The 
present study has found that the trunk and pelvis remain more in-phase as velocity (normalized 
stride length and cadence) increased.  This more in-phase relationship also supports the 
supposition that the body is partially using a forceful toe-off to meet the demands of the imposed 
velocity rather than additional pelvic rotation a velocities above Normal, which would cause the 
upper body to counter rotate to conserve total body angular momentum. 
Section 2: Walking Cadence Effects 
The purpose of this section of the study was to investigate the effect of altered cadence on 
spatiotemporal variables, relative phase differences, and electromyography at three levels 
(Normal, 125% of Normal, 75% of Normal cadence). The kinematic variables of interest were 
height normalized stride length (SL), walking velocity, relative phase mean differences for the 
trunk-pelvis, trunk-leg, and pelvis leg (TP mean, TL mean, PL mean, respectively), relative phase 
standard deviation differences for the trunk-pelvis, trunk-leg, and pelvis-leg (TP sd, TL sd, PL sd, 
respectively), and angular positions of the pelvis, trunk, and leg at foot strike. Table 16 illustrates 
the spatiotemporal variables and relative phase differences for the altered cadence conditions as 
compared to the normal cadence condition.  
 
 
122 
 
Table 16 - Effect of Cadence on all Kinematic Variables as 
Compared to Normal 
Variable 125% 75% 
SL ? ? 
Velocity ? ? 
TP mean ? ? 
TP sd ? ? 
TL mean ? ? 
TL sd ? ? 
PL mean ? ? 
PL sd ? ? 
? pelvis ? ? 
? trunk ? ? 
? leg ? ? 
Note.  ? = no significant difference, ? = increase, ? = decrease 
 
All indicated differences were significantly different (p <0.05) 
 
The sEMG variables of interest were the normalized peak activity during the stance phase 
of the peroneus longus (PerL peak), soleus (Sol peak), biceps femoris (BF peak), right gluteus medius 
(rGM peak), left gluteus medius (lGM peak), and erector spinae (ES peak). Furthermore, the time to 
onset muscle activity for the same six muscles at foot strike (PerL onset, Sol onset, BF onset, rGM 
onset, lGM onset, and ES onset) were also investigated. Table 17 illustrates the sEMG (normalized 
peak and time to onset for the respective six muscles) differences for the altered cadence 
conditions as compared to the normal cadence condition.  
 
 
 
 
123 
 
Table 17 - Effect of Cadence on all sEMG Variables as 
Compared to Normal 
Variable 125% 75% 
PerL peak  ? ? 
Sol peak  ? ? 
BF peak  ? ? 
rGM peak  ? ? 
lGM peak  ? ? 
ES peak  ? ? 
PerL onset  ? ? 
Sol onset  ? ? 
BF onset  ? ? 
rGM onset  ? ? 
lGM onset  ? ? 
ES onset  ? ? 
Note.  ? = no significant difference, ? = increase, ? = decrease 
 
All indicated differences were significantly different (p <0.05) 
 
Condition Differences in Spatiotemporal & Kinematic Variables During Altered Cadence 
The present study found a significant difference in velocity between all cadence 
conditions (Figure 28), indicating that as velocity was increased from 75% of Normal, to 
Normal, to 125% of Normal, participants adjusted velocity by continuing to walk faster. As with 
the altered velocity conditions, significant differences in height normalized stride length between 
cadence conditions (Figure 29) were found, indicating that as cadence was increased from 75% 
of Normal, to Normal, to 125% of Normal, participants adjusted stride length by taking longer 
height normalized strides. Similar to when velocity was altered, the present section demonstrated 
significant increases in PL peak (Figure 37) and Soleus peak (Figure 38) as cadence increased. A 
124 
 
similar hypothesis as proposed to explain the changes noted during the imposed velocity 
condition is believed to have contributed to the significant difference in the behavior of the 
peroneus longus and soleus during the stance phase. It is believed that the significant increases in 
the peroneus longus, acting as a plantarflexor and evertor of the foot during the stance phase, and 
the soleus, performing plantarflexion of the foot during the stance phase, assisted in an increase 
in propulsion and as a result a greater excursion of the center of mass. Specifically, it is believed 
the participant would need to rapidly evert the foot to better load the head of the metatarsals (e.g. 
ball of the foot) and the hallux (e.g. big toe) during toe-off.   As velocity (and cadence) 
increased, the rate of eversion would occur faster due to shorter absolute ground contact times, 
thereby explaining the increases in peroneus longus activity as the higher cadence levels. 
Similarly, the strong plantarflexion moment created by the soleus during the stance phase of gait, 
particularly terminal stance, would assist in the center of mass excursion.  
Tables 18, 19, and 20 display the height normalized stride length, velocity, and cadence, 
respectively for the altered velocity and altered cadence conditions. Though differences in the 
dependent variables between conditions (i.e. velocity vs. cadence) were not tested for 
significance, values are displayed for comparison purposes. For instance, while normalized stride 
lengths were almost equal during self-selected trials (Normal), differences were observed at the 
75% and 125% trials, with an almost 2% difference in normalized stride length between 125% of 
Normal Velocity and 125% of Normal Cadence (Table 18). 
 
 
 
 
125 
 
Table 18 - Changes in Normalized Stride Length 
During Altered Velocity & Altered Cadence 
  Velocity Conditions Cadence Conditions 
125% of Normal 90.5% 88.6% 
Normal 80.7% 80.8% 
75% of Normal 71.3% 70.6% 
Note. Reported values are a percentage of standing 
height 
 
When considering the preferred walking velocity (Normal) of the participants, velocity 
was once again almost identical between testing sessions; however, large differences were found 
between the 125% and 75% conditions, with the largest differences exhibited between 125% of 
Normal Velocity and 125% of Normal Cadence (Table 19). Similar to the findings for 
normalized stride length, this suggests participants reacted differently when participants were 
free to choose a cadence and stride length combination (i.e. altered velocity) to achieve the 
required velocity than when cadence was pre-determined. Furthermore, when cadence was 
altered from 75% of Normal to 125% of Normal, a considerably greater range (from 75% of 
Normal to 125% of Normal) is noted. 
Table 19 - Changes in Velocity During Altered Velocity 
& Altered Cadence 
  Velocity Conditions Cadence Conditions 
125% of Normal 1.59 1.76 
Normal 1.28 1.30 
75% of Normal 0.96 0.87 
Note. Reported values are in ms-1 
 
 
 As with velocity and cadence, the preferred walking cadence (Normal) was similar 
between testing sessions, large differences were found between the 125% and 75% conditions, 
with the largest differences exhibited between 125% of Normal Velocity and 125% of Normal 
Cadence (Table 20). Rather than adjusting stride length to a greater extent to adjust for the larger 
126 
 
variation in cadence during the cadence conditions, participants appeared to not adjust stride 
length outside of an internally selected range, creating larger fluctuations in velocity (Table 19) 
than was observed during the altered velocity conditions; however, a limit may have been 
reached at the upper limits (125% of Normal Cadence) as participants displayed a 2% lower 
normalized stride length during the faster condition (Table 18).  
Table 20 - Changes in Cadence During Altered Velocity 
& Altered Cadence 
  Velocity Conditions Cadence Conditions 
125% of Normal 122.9 139.6 
Normal 111.3 112.5 
75% of Normal 93.7 84.7 
Note. Reported values are in steps*minute-1 
 
It is hypothesized that the dual-task nature of walking and coordinating walking patterns 
to a metronome altered walking patterns as compared to the single-task nature of the altered 
velocity conditions. While the altered cadence condition was successful in achieving the 125% 
and 75% of Normal cadences, greater variability in the kinematics of the trunk, pelvis, and right 
thigh at foot strike were noted. As opposed to the altered velocity conditions, significant main 
effects were found during the altered cadence for all six kinematic variables of interest. For 
instance, during the altered cadence condition, the trunk and pelvis were found to be most 
synchronous (Figure 30) and least variable (Figure 31) during the 125% of Normal condition. 
Conversely, the relationship between the trunk and leg (Figure 32), as well as the pelvis and leg 
(Figure 33), indicated greater asynchronicity during the 125% of Normal and 75% of Normal 
conditions as compared to Normal. Therefore, it is hypothesized that the increasing velocity that 
occurred as cadence was increased, and accompanied increases in normalized stride length, 
altered the relationship between the trunk and pelvis during overground walking. Conversely, the 
127 
 
larger variation in cadence that occurred during the altered cadence conditions altered the 
relationship between the trunk and leg as well as the pelvis and leg.  
These findings are interesting in light of the of the other kinematic results which 
indicated that the pelvis exhibited right pelvic girdle rotation and left trunk rotation at foot strike, 
as opposed to left pelvic girdle rotation (Figure 34) and right trunk rotation (Figure 35) as during 
the Normal and 75% conditions. It is believed that the dual-task activity of walking to a 
metronome, which concurrently fixed a primary spatiotemporal variable, significantly altered the 
timing of gait events and joint kinematics. For instance, no metronome was utilized during the 
Normal conditions as these trials served as baseline trials. Therefore, the only trials in which the 
metronome was utilized was during the 125% of Normal and 75% of Normal cadence 
conditions. Conversely, results of the present section indicated that upper leg displayed greater 
external rotation at foot strike during the 125% of Normal cadence condition as compared to 
Normal and 75% of Normal (Figure 36), a finding similar to the altered velocity section. It is 
believed that the altered trunk and pelvis positions at foot strike necessitated considerably larger 
external rotation of the upper leg during the higher cadence level due to the increased velocity 
and narrower step width which occurs as velocity increases (Perry, 1992; Whittle, 2003).  
The present study also found a significant increase in LPH musculature (e.g. biceps 
femoris, bilateral gluteus medius, erector spinae) activity during the 125% of Normal conditions 
relative to other conditions as cadence was increased (Figure?s 39-42). This finding is similar to 
the increased LPH that was observed as velocity increased as discussed in Section 1. It is 
believed that the increased LPH activity may have arisen from two separate areas. The primary 
hypothesis is that as stride length and cadence increased, step width (not reported) would have 
also decreased, placing additional demands on the respective gluteus medius to assist in 
128 
 
maintaining frontal plane pelvic neutrality during single-limb support (Perry, 1992; Whittle, 
2003). Support for this theory also comes from the significantly altered relationship between the 
pelvis and leg during the altered cadence conditions relative to Normal. Increased gluteus medius 
and erector spinae activity may have also resulted from phase-locking of the trunk and pelvis, 
which may have resulted in increased co-contraction of LPH musculature. Phase locking of the 
pelvis has been previously hypothesized to occur during overground walking at greater than 
normal velocities (Crosbie & Vachalathiti, 1997).  
Section 3: Footwear Effects During Walking 
 Previous research on relative phase differences has primarily focused on changes in 
synchronicity as a result of alterations in velocity, the author is unaware of any study that 
investigated the effect of footwear on these variables (Bruijn et al., 2008; Y. Huang et al., 2010; 
C. J. Lamoth, Beek, et al., 2002; van Emmerik & Wagenaar, 1996). This may be important due 
to recent research highlights of significant mechanical differences in shod gait relative to 
barefoot walking (Hamill, Russell, Gruber, & Miller, 2011; Lieberman et al., 2010; Majumdar, 
Banerjee, Pal, Kumar, & Selvamurthy, 2006; Morio et al., 2009; Wegener et al., 2011). Though 
the reasons for these differences are not entirely understood, the literature appears to support a 
neural interaction (i.e. reduced tactile feedback) rather than a purely mechanical (i.e. increased 
mass of the distal segment) (Divert et al., 2005; Divert et al., 2008; Maki et al., 1999; Manor & 
Li, 2009; Nurse et al., 2005; Nurse & Nigg, 2001). The mass of the footwear utilized in the 
present study was an average per foot of 143g for SH, 162g for INSH, and 31.5g for IN. While 
the shod conditions (SH, INSH) were similar to the lightest footwear option utilized by Divert 
and colleagues (2008), the insole only condition was considerably lighter than anything presently 
reported in the literature.  
129 
 
Previous research has suggested it may be possible to alter the behavior of the 
musculature through the manipulation of footwear components or enhancement of tactile 
feedback (Murley et al., 2009; Nurse et al., 2005; Nurse & Nigg, 2001; Scott et al., 2012). 
Therefore, the purpose of this portion of the study was to investigate the effect of footwear on 
spatiotemporal variables, relative phase differences, and muscle activity during walking. Normal 
walking velocity and cadence were chosen for four randomized footwear conditions (SH, BaFt, 
IN, INSH) to examine differences in the chosen dependent variables. The kinematic variables of 
interest were height normalized stride length (SL), cadence, walking velocity, relative phase 
mean differences for the trunk-pelvis, trunk-leg, and pelvis leg (TP mean, TL mean, PL mean, 
respectively), relative phase standard deviation differences for the trunk-pelvis, trunk-leg, and 
pelvis-leg (TP sd, TL sd, PL sd, respectively), and angular positions of the pelvis, trunk, and leg at 
foot strike. Table 21 illustrates the spatiotemporal variables and relative phase differences as 
compared to the SH condition. 
 
 
 
 
 
 
 
 
 
 
130 
 
Table 21 - Effect of Footwear on all Kinematic 
Variables as Compared to Shod   
Variable BF IN INSH 
SL ? ? ? 
Cadence ? ? ? 
Velocity ? ? ? 
TP mean ? ? ? 
TP sd ? ? ? 
TL mean ? ? ? 
TL sd ? ? ? 
PL mean ? ? ? 
PL sd ? ? ? 
? pelvis ? ? ? 
? thorax ? ? ? 
? leg ? ? ? 
Note.  ? = no significant difference, ? = increase, ? = decrease 
 
All indicated differences were significantly different (p <0.05) 
 
 
The sEMG variables of interest was the normalized peak activity during the stance phase 
of the peroneus longus (PerL peak), soleus (Sol peak), biceps femoris (BF peak), right gluteus medius 
(rGM peak), left gluteus medius (lGM peak), and erector spinae (ES peak). Furthermore, the time to 
onset muscle activity for the same six muscles at foot strike (PerL onset, Sol onset, BF onset, rGM 
onset, lGM onset, and ES onset) were also considered. Table 22 illustrates the sEMG (normalized 
peak and time to onset) differences as compared to the SH condition. 
 
 
 
 
 
131 
 
Table 22 - Effect of Footwear on all sEMG 
Variables as Compared to Shod   
Variable BF IN INSH 
PerL peak  ? ? ? 
Sol peak  ? ? ? 
BF peak  ? ? ? 
rGM peak  ? ? ? 
lGM peak  ? ? ? 
ES peak  ? ? ? 
PerL onset  ? ? ? 
Sol onset  ? ? ? 
BF onset  ? ? ? 
rGM onset  ? ? ? 
lGM onset  ? ? ? 
ES onset  ? ? ? 
Note.  ? = no significant difference, ? = increase, ? = decrease 
 
All indicated differences were significantly different (p <0.05) 
 
Footwear effects on Spatiotemporal & Kinematic Variables 
 While footwear was found to significantly affect walking velocity, with BaFt and IN 
conditions exhibiting a significantly lower velocity relative to shod matched condition (i.e. SH, 
INSH), this difference may be of minimal clinical or sport application has noted by the small 
effect size (Figure 44). Conversely, height normalized stride length was found to be significantly 
shorter, and cadence significantly higher, during the BaFt and IN condition as compared to SH 
and INSH. While the difference in these spatiotemporal parameters for barefoot and shod 
walking is supported by previous research, previous research has indicated enhanced tactile 
feedback to significantly alter postural stability and muscle activity during standing and 
overground walking (Majumdar et al., 2006; Maki et al., 1999; Nurse & Nigg, 2001; Romer et 
al., 2012; Wegener et al., 2011). The lack of a significant difference in normalized stride length 
between insole matched conditions (BaFt vs. IN; SH vs. INSH) appears to lend support for a 
132 
 
possible inertial interaction of footwear on gait mechanics rather than reduced afferent feedback 
as the mass of the insole was negligible in the current design. Likewise, the significantly higher 
cadence during the BaFt conditions relative to the IN condition lends support for a mass effect of 
footwear, though admittedly the mass of the IN was minimal (31.5g).  
 Though significant footwear effects were noted in the walking velocity conditions, the 
small changes in velocity were a result of a significantly shorter normalized stride length of the 
BaFt and IN conditions which were accompanied by slight, significant increases in cadence. 
Wegener et al. (2011) reported similar differences in spatiotemporal variables (e.g. velocity, 
cadence, stride length) when examining barefoot vs. shod gait, while also reporting significant 
reductions in base of support, contact time, and double support time during barefoot walking. 
Therefore, it is believed the BaFt and IN conditions resulted in shorter relative stance phase and 
reduced ground contact time. As during the altered cadence conditions, which resulted in 
increased pelvis-leg asynchronicity at higher cadence levels, it is hypothesized that the altered 
ground contact time and cadence of the BaFt and IN conditions resulted in a significant 
difference in synchronicity between the pelvis and leg.  
 While footwear was found to not significantly alter the kinematics of any of the other 
segments investigated or the sEMG of the muscles investigated, the present section has advanced 
this area of research by examining the effects of footwear on relative phase differences and 
segment kinematics during normal, overground walking. The present study indicated that while 
footwear did not alter the level of synchronicity between the trunk-pelvis and trunk-leg, footwear 
did alter the level of pelvis-leg synchronicity. Furthermore, no differences in trunk, pelvis, or leg 
rotation were noted as a result of footwear, which may indicate that footwear does not have a 
significant effect on normal, overground walking. Instead, footwear effects on joint kinematics at 
133 
 
foot strike may be more primary constrained to differences in hip, knee, and ankle sagittal plane 
kinematics which were not investigated in the present study.  
 Conversely, the shoe utilized for the present study (Adidas AdiPure Adapt, Portland, OR, 
USA) was of a minimalist, zero heel height differential which were designed by Adidas to 
resemble a bare foot. Therefore, the shoes utilized in the present study are dissimilar from most 
commercially available footwear, of which typically relies on thicker sole material, considerable 
heel height differentials, and relatively inflexible midsole material which reduces natural foot 
motion through the gait cycle. Though the Adidas AdiPure Adapt were chosen specifically to 
reduce the effect of typical footwear design on the dependent variables, additional investigation 
is needed to understand the effects of heel height differentials and midsole flexibility on joint 
kinematics and muscle activity during walking.  
Gender Differences in Spatiotemporal & Kinematic Variables 
 Similar to the previous section which examined footwear effects, gender differences were 
noted when examining the effect of altered footwear during normal walking. Specifically, the 
present study found a significant gender difference for height normalized stride length when 
footwear was altered (Figure 46) with females exhibiting a significantly longer normalized stride 
length than males. Furthermore, the difference in normalized stride length between genders 
equaled approximately 3.5% of standing height. Therefore, it appears that females take longer 
height normalized strides than males during self-selected walking. The lack of a significant 
gender effect for trunk or pelvis joint kinematics suggests that females achieved the greater stride 
length through increased center of mass excursion, though no significant gender differences were 
noted for the sEMG of any of the muscles. 
 Significant gender differences for cadence (Figure 48) were also found, with females 
134 
 
exhibiting a significantly higher cadence than males. Specifically, the difference in cadence 
between genders was approximately 8.4 steps*min-1. Finally, females walked slightly, faster 
(1.29 ms-1) than males (1.25 ms-1); however, the difference in walking velocity was exceptionally 
small and was non-significant. Rather, it is believed that the significant difference in cadence is a 
result of more profound differences, particularly differences in anthropometrics (e.g. shorter leg 
length of females). 
 Results from the present study indicated a significant gender difference for TP SD when 
footwear was altered (Figure 49), with males exhibiting a greater TP SD, indicating males 
exhibited greater variability in the relationship between trunk and pelvis movements. Though TP 
mean approached significance, with males exhibiting greater asynchronicity than females, this 
value did fail to reach the apriori alpha level. Nonetheless, the results of the present study 
suggest that males may be more variable than females in the movement of the LPH complex 
during normal walking.  
Section 4: Summary of the Discussion 
 
 The purpose of this section is to summarize the findings of the present study while 
addressing the individual research questions. This summary includes three parts: 
? Part 1: Summarize the discussion comparing the effects of altering velocity and 
cadence on spatiotemporal variables, relative phase measures, and sEMG. 
? Part 2: Summarize the discussion comparing the effects of footwear on the 
spatiotemporal variables, relative phase measures, and sEMG. 
? Part 3: Summarize the discussion of gender differences in spatiotemporal 
variables and relative phase measures. 
 
135 
 
Summary: Part 1 
Spatiotemporal & Kinematic Variables 
As velocity and cadence were controlled at 75% of Normal, Normal, and 125% of 
Normal there was a significant increase in the normalized stride length of the participants. 
Increases in stride length as velocity increased was not necessarily an unexpected result given 
that walking velocity is the product of stride length and cadence (Whittle, 2003; Wyke, 1967). 
Therefore, concurrent increases in stride length and cadence would be expected as velocity is 
increased (Blanc et al., 1999; Perry, 1992). For instance, van Emmerik and Wagenaar (1996) 
found an increase in stride length as walking velocity was increased from 0.3 ms-1 to 1.3ms-1. 
Likewise, Crosbie et al. (1997b) found significantly longer height normalized step lengths when 
comparing freely chosen velocities at two levels: Normal and Faster than Normal (participants 
were instructed to walk as if they were late to a meeting).  
In the present study, both velocity and cadence increased when manipulated from 75% of 
Normal to 125% of Normal, indicating that the author was successful in manipulating the 
independent variables for each session. However, both velocity and cadence increased 
significantly on the sessions in which it was a dependent variable as well. The latter results may 
have confounded the overall results, though it would have likely been impossible to 
simultaneously control for both spatiotemporal variables during overground walking. Therefore 
this is a limitation of the present study.   
When examining the effect of altered velocity or altered cadence, significant differences 
were found for TP mean, TP SD, TL mean, TL SD, PL mean, PL SD, Pelvic Rotation, Trunk Rotation, 
and Leg Rotation; however the influence of velocity appeared to be smaller in which significant 
differences were found only for TP mean, Pelvic Rotation, and Trunk Rotation. Furthermore, 
136 
 
findings from the present study appear to contradict the findings of earlier studies, which found a 
decrease in the TP mean, or increase in the synchronicity of the trunk and pelvis, as velocity was 
increased. Of noteworthy importance though is the considerable difference in velocity ranges of 
the present study as compared to previous studies. For instance, van Emmerik and Wagennar?s 
(1996) had participants walk on a treadmill at a pre-determined velocity, beginning at 0.3 ms-1 
and ascending to 1.3 ms-1 in 0.1 ms-1 intervals before decreasing back to the original velocity. 
Conversely, the participants in the present study exhibited an average walking velocity between 
0.87 ms-1 and 1.76 ms-1. Interestingly, the maximum walking velocity in the van Emmerick and 
Wagennar study (1996) was approximately the same as the Normal conditions in the present 
study. Lamoth, Beek, and Meijer (2002) published a similar paper examining trunk-pelvic 
coordination in the transverse plane during gait. Beginning at 0.39 ms-1, walking velocity was 
increased in 0.22 ms-1 until reaching a maximum of velocity of 1.5 ms-1, the authors reported a 
significant increase in the TP mean as velocity increased, contradicting results of the present study. 
Again, though, both Lamoth et al. (2002) and van Emmerick and Wagennar (1996) performed 
their respective studies on treadmills using relatively rudimentary motion capture technology as 
compared to the present study.  
Another component that may explain the differences in results is the methodological 
approach used in the present study, which relied on a randomization of high (125%) and low 
(75%) velocities and cadences based on the a participants self-selected gait patterns. Therefore, 
velocities were not pre-determined, nor were they altered in a predictable format as was 
performed by almost all of the previous researchers (Bruijn et al., 2008; Y. Huang et al., 2010; C. 
J. Lamoth, Beek, et al., 2002; Wagenaar & van Emmerik, 2000). Another possibility is the 
difference in kinematic collections between past research, which primarily relied on LED or 
137 
 
selspot style systems, whereas the present project utilized a passive motion capture system and a 
marker set consisting of 44 markers (Table 2). Finally, while many of the previous articles 
utilized a discrete Fourier transform or non-normalized data in the construction of the relative 
phase plots, the present study normalized angular displacements and velocities prior to the 
construction of all relative phase plots (Bruijn et al., 2008; Y. Huang et al., 2010; C. J. Lamoth, 
Beek, et al., 2002). Kurz & Stergiou (2002) reported that the normalization of phase angle 
calculations resulted in significant and profound differences in the determination of the relative 
phase values when compared to non-normalized data. Specifically, the authors reported RMS 
differences of between 22.7? to 70.6? when comparing non-normalized relative phase angles to 
normalized phase angles (Kurz & Stergiou, 2002). Additional investigations should be 
completed utilizing a standardized data reduction process, such as the one utilized in the present 
study, so that results are able to be more equally compared and discussed.  
Finally, previous research has relied exclusively on motorized treadmills to manipulate 
walking velocities. Furthermore, these studies have manipulated walking velocity without 
reference to a participant?s preferred walking strategy. As previously stated, several studies have 
noted significant biomechanical differences, particularly at the knee and hip, in treadmill walking 
relative to overground walking (Alton et al., 1998; Dingwell et al., 2001; S. J. Lee & Hidler, 
2008; Parvataneni et al., 2009). Furthermore, previous research has demonstrated that treadmill 
walking resulted in changes to both variability and local stability relative to overground walking 
(Dingwell et al., 2001). Another importance difference in overground walking has been 
previously noted by Crosbie and Vachalathiti (1997), which reported that walking at velocities 
above normal tended to produce a tighter phase locking of movement patterns. Therefore, if the 
previous research failed to adequately manipulate the velocities of the participants, as the 
138 
 
references were absolute and not scaled to the individual, the maximal treadmill walking speed 
may have barely reached normal walking speeds for the participants. 
Surface Electromyography Variables 
As hypothesized, normalized peak sEMG during the stance phase increased as velocity 
and cadence increased from 75% of Normal, to Normal, to 125% of Normal for most muscles. 
Interestingly, from the present study were the lack of significant difference between many of the 
75% of Normal and Normal conditions, specifically the soleus during the altered velocity 
conditions, and the peroneous longus, biceps femoris, bilateral gluteus medius, and erector 
spinae during the altered cadence conditions. The results may indicate the level of muscle 
activity required during the gait cycle, from weight acceptance to toe-off, is not significantly 
different during the lower conditions (75%, Normal) due to a minimum amount of muscle 
activity required to perform the requisite tasks. Conversely, the dual-task of the altered cadence 
conditions may have significantly altered single and support times, as suggested by Tables 18 ? 
20, altering muscle activity during 75% cadence trials.  
It is believed the increase in biceps femoris activity and erector spinae activity during the 
stance phase is partially due to the increased center of mass excursion that is occurring at higher 
velocities, chiefly an increase in normalized stride length (Table 18). Conversely, the increase in 
the muscle activity for the aforementioned multi-articular muscles may be due to the significant 
alterations in segment synchronization and rotational amplitude as velocity and cadence 
increased. Specifically, though rotational amplitudes for the trunk and pelvis tended to increase 
as velocity and cadence increased, there was also a corresponding increase in the synchronization 
of the trunk-pelvis. It is believed that this increase synchronization may have occurred through a 
type of phase locking, which Crosbie and Vachalathiti (1997) has hypothesized to occur at 
139 
 
greater than normal walking velocities. The present author asserts that this phase locking may be 
increasing the co-contraction of the musculature of the LPH complex. 
Finally, no significant difference was noted in the onset of muscle activity for any of the 
six muscles observed in the present study. Future research may want to examine the time-to-peak 
muscle activity, which may be more sensitive to variation in loading rate at foot strike when 
velocity and cadence are altered.  
Summary: Part 2 
Spatiotemporal & Kinematic Variables 
 Results indicated that footwear had a significant effect on the preferred walking velocity, 
cadence, and normalized stride length of the participants. Specifically, the barefoot and barefoot-
like condition (IN) exhibited a significantly shorter normalized stride length relative to both SH 
and INSH. Furthermore, the matched footwear conditions (BaFt and IN, SH and INSH), 
exhibited no significant difference between each other, indicating the textured insole had no 
significant effect on normalized stride length. Cadence was greatest for the BaFt condition, 
which exhibited significant differences from SH, IN, and INSH. Finally velocity was lowest for 
the barefoot and barefoot-like condition (IN), as compared to both SH and INSH. Results 
therefore would appear to support the increased mass of the distal segment when wearing shoes 
is the cause for an increased stride length relative to barefoot (Majumdar et al., 2006); however, 
other studies appear to discount this hypothesis (Divert et al., 2005; Divert et al., 2008; Maki et 
al., 1999; Nurse et al., 2005; Shroyer & Weimar, 2010). While the mass of the shod conditions in 
the present study were minimal (SH = 143g, INSH = 162g), even this slight difference may be 
enough to result in alterations in the timing of key gait events, particularly foot strike. Future 
research should be expanded to include greater manipulation of the mass of a shoe, variation in 
140 
 
texture designs, and the effect of heel height differentials. Additionally, future research and 
analyses should examine ground reaction forces (e.g. impact transients) as well as lower 
extremity joint kinematics as the use of stride length alone may be hiding true mass or sensory 
effects.  
 The only kinematic variable to display a significant footwear effect was PL mean with 
BaFt and IN exhibiting greater asynchronicity than the SH and INSH conditions. Though the 
present study is the first known study to examine the footwear effects on relative phase 
differences, the results appear to support research by Huang and colleagues (2010), which found 
that participants exhibited a more-in phase PL mean when participants took significantly longer 
strides, with a concurrent reduction in cadence, while walking at a fixed velocity on a treadmill. 
Therefore, results from the present study suggest that an increase in stride length and reduction in 
cadence will create a more in-phase pelvis-leg relationship. 
Surface Electromyography Variables 
 Interestingly, no significant effects were found for normalized peak muscle or onset of 
muscle activity for any of the six muscles observed during the study. The lack of significant 
difference in normalized peak muscle activities was unexpected considering previous researchers 
have noted significant footwear effects on peak muscle activity (Bird et al., 2003; Gefen et al., 
2002; Li & Hong, 2007; Murley et al., 2009; Nurse & Nigg, 2001). The results may have been 
confounded by the minimalist nature of the footwear design, which is extremely flexible, light, 
and marketed as a minimalist running shoe with a zero heel height differential.  
 Finally, no significant effect was found for footwear for the time to onset of muscle 
activity for any of the muscles observed. Nurse et al. (2005) has previously shown a significant 
difference in the time to onset when comparing a smooth vs. textured insole during walking, 
141 
 
however, the authors in that study relied on a wavelet analysis of the EMG signal (Nurse et al., 
2005). Conversely, the present studied maintain the EMG signal in the temporal component, 
utilizing a double-threshold to determine onset. As suggested by Murley et al. (2009), it is 
difficult to compare findings from wavelet analyses to traditional EMG analysis. However, 
several other authors have noted a significant effect of increased afferent feedback in altering 
postural or gait kinematics  (Bird et al., 2003; Kelleher et al., 2010; Maki et al., 1999; Nurse et 
al., 2005; Nurse & Nigg, 2001). For instance, Maki and colleagues (1999) reported that the 
enhanced plantar facilitation on the periphery of the foot reduced COP trajectory and time to 
stabilization in response to unpredicted postural perturbations. In a study utilizing the same 
textured insoles as the present study, Hatton et al. (2011) reported a 9.2% reduction in 
mediolateral sway in older adults during a static, eyes closed test. Future research should be 
directed towards texture variation as well as the examination of adaptation periods to better 
understand the augmented afferent feedback on postural stability, gait kinematics, and muscle 
activity. 
Summary: Part 3 
 Spatiotemporal & Kinematic Variables 
 In the present study, females exhibited a significantly longer normalized stride length 
than males for the altered footwear condition only. Though unusual, Crosbie and colleagues 
(1997b) reported a significant gender difference in height normalized step length during a faster 
than normal walking condition, with females exhibiting a significantly longer step than males. 
Similarly, females were found to display a significantly higher cadence than males altered 
footwear conditions. Noteworthy though is the lack of a significant difference in walking 
velocity, suggesting that the significantly longer normalized strides of females may be resulting 
142 
 
from differences in center of mass excursion techniques. Likewise, as discussed in Section 3, the 
difference in normalized stride length was small, with a small effect size, and moderate power.  
 Relative phase differences were only found to be significantly different during the 
footwear conditions, with males exhibiting significantly greater TP SD indicating that the trunk 
and pelvis of males was significantly more variable than for females when footwear was altered. 
As previously stated, the findings are especially interesting considering that women exhibited a 
greater normalized stride length, which has been previously hypothesized to result in a more-out-
of phase relationship between the trunk and pelvis; however, all of those previous research 
studies have been conducted utilizing various differences in methodology including treadmill 
walking and velocities not scaled to the preferred spatiotemporal variables (Bruijn et al., 2008; 
Y. Huang et al., 2010; C. J. Lamoth, Beek, et al., 2002; van Emmerik & Wagenaar, 1996). 
Finally, no significant gender differences were found for the angular position of the trunk, pelvis, 
or leg at foot strike.  
The present study is the first known study to examine gender differences in the 
aforementioned kinematic variables. Unique to the present study is the finding that healthy males 
tend to display greater asynchronicity between the trunk and pelvis, which may be related to an 
increased variability of the pelvis during overground walking, though the source for these gender 
differences are not yet understood. Future research should be directed towards understanding if 
this relationship is similar across various footwear designs and age groups (i.e. adolescents, older 
adults, elderly). Additionally, future research should be directed towards studying the source of 
the gender differences in the relative phase measures, which may aid in understanding the 
clinical relevance of the measures. 
 
143 
 
Section 5: Future Research 
Future Research for Velocity & Cadence Effects 
When examining relative phase differences, the present focus within the literature has 
been on the influence of velocity on the relationship between trunk, pelvis, and leg movements 
(Bruijn et al., 2008; Y. Huang et al., 2010; C. J. Lamoth, Beek, et al., 2002; van Emmerik & 
Wagenaar, 1996). Though it has long been hypothesized that longer stride lengths would result in 
more asynchronous movements, only recently have researchers attempted to separate the 
variables that compose velocity (Y. Huang et al., 2010). Similarly, the present project 
manipulated velocity and cadence during a previously unstudied task, overground walking, in an 
attempt to gain a better understanding of the segmental rotations during walking. While stride 
length does appear to be associated with an altered trunk-pelvis relationship, cadence also 
appears to be an important factor in the trunk-leg and pelvis-leg movements. Further researcher 
should be directed towards the manipulation of stride length and cadence to better gauge the 
independent contributions of these variables to segmental synchronicity. Additionally, future 
research should be directed towards gaining a better understanding of the clinical relevance of 
the relative phase measures. Finally, there is a need to develop an improved methodological 
approach for the manipulation of velocity, cadence, and stride length, thereby minimizing the 
cognitive demands possibly associated with altering spatiotemporal variables during overground 
walking. 
Future Research for Footwear Effects 
 The present study was one of the first to research the effect of footwear on relative phase 
differences, during treadmill or overground walking. Minimal differences were found except for 
spatiotemporal differences which have generally been accepted within the literature (Bird et al., 
144 
 
2003; Lieberman et al., 2010; Morio et al., 2009; Murley et al., 2009; Sato et al., 2012; Scott et 
al., 2012; Wegener et al., 2011). In an attempt to isolate only the influence of the foot in a shoe 
and not characteristics of the shoe, a minimalist, zero-drop shoe was chosen.  Unfortunately, this 
footwear condition may not have had the necessary inertia or sole rigidity to produce previously 
noted footwear differences. On the contrary, it may be that while shoes result in significant 
differences in lower extremity and some muscle activity, these differences are not realized in a 
variable such as relative phase measures or time to onset of muscle activity. Though the shoe 
chosen for the present study was identified for the minimalist nature, constraints to use for the 
general population or many occupational tasks are contraindicated. Therefore, many questions 
remain as to whether a more traditional shoe design will significantly alter the relationship 
between trunk, pelvic, and hip movements or the efferent response of the lower extremity.  
Future Research for Gender Differences 
 Interestingly, males and females appeared to display the greatest differences when 
footwear effects were examined. Specifically, males appeared to be more asynchronous and 
variable in trunk and pelvic. Past research has primarily focused on pathological populations, 
such as individuals with low back pain, or females with pelvic girdle related pain (Crosbie & 
Vachalathiti, 1997; Crosbie et al., 1997a; Y. Huang et al., 2010; Y. P. Huang et al., 2011; C. J. 
Lamoth, Meijer, et al., 2002; W. Wu et al., 2002; W. Wu et al., 2004; W. H. Wu et al., 2008). In 
the present study, though females displayed a longer normalized stride length and reduced 
variability in trunk-pelvic movements. Therefore future research should be directed towards 
gaining a better understanding of the gender differences and the significance of the differences in 
healthy populations before attempting to extrapolate to clinical populations.  
 
 
145 
 
References 
 
Aghazadeh, F., & Lu, H. (1994). Relationship between Posture and Lifting Capacity. 
International Journal of Industrial Ergonomics, 13(4), 353-356.  
 
Allet, L., Armand, S., de Bie, R. A., Pataky, Z., Aminian, K., Herrmann, F. R., & de Bruin, E. D. 
(2009). Gait alterations of diabetic patients while walking on different surfaces. 
[Comparative Study]. Gait Posture, 29(3), 488-493. doi: 10.1016/j.gaitpost.2008.11.012 
 
Alton, F., Baldey, L., Caplan, S., & Morrissey, M. C. (1998). A kinematic comparison of 
overground and treadmill walking. Clinical Biomechanics, 13(6), 434-440. doi: 
http://dx.doi.org/10.1016/S0268-0033(98)00012-6 
 
Andersson, G. B. (1998). Epidemiology of low back pain. Acta Orthop Scand Suppl, 281, 28-31.  
 
Aultman, C. D., Drake, J. D., Callaghan, J. P., & McGill, S. M. (2004). The effect of static 
torsion on the compressive strength of the spine: an in vitro analysis using a porcine spine 
model. [Comparative Study Research Support, Non-U.S. Gov't]. Spine (Phila Pa 1976), 
29(15), E304-309.  
 
Aultman, C. D., Scannell, J., & McGill, S. M. (2005). The direction of progressive herniation in 
porcine spine motion segments is influenced by the orientation of the bending axis. [In 
Vitro Research Support, Non-U.S. Gov't]. Clin Biomech (Bristol, Avon), 20(2), 126-129. 
doi: 10.1016/j.clinbiomech.2004.09.010 
 
Ayyappa, E. (1997). Normal Human Locomotion, Part 1: Basic Concepts and Terminology  
Retrieved January 1, 2013, from 
http://www.oandp.org/jpo/library/1997_01_010.asp?searchquery=gait%20ayyappa?searc
hquery=determinants 
 
Bird, A. R., Bendrups, A. P., & Payne, C. B. (2003). The effect of foot wedging on 
electromyographic activity in the erector spinae and gluteus medius muscles during 
walking. Gait Posture, 18(2), 81-91.  
 
Blanc, Y., Balmer, C., Landis, T., & Vingerhoets, F. (1999). Temporal parameters and patterns 
of the foot roll over during walking: normative data for healthy adults. [Research 
Support, Non-U.S. Gov't]. Gait Posture, 10(2), 97-108.  
 
Bramble, D. M., & Lieberman, D. E. (2004). Endurance running and the evolution of Homo. 
[Historical Article Research Support, Non-U.S. Gov't]. Nature, 432(7015), 345-352. doi: 
10.1038/nature03052 
 
Braunstein, B., Arampatzis, A., Eysel, P., & Bruggemann, G. P. (2010). Footwear affects the 
gearing at the ankle and knee joints during running. [Comparative Study]. J Biomech, 
43(11), 2120-2125. doi: 10.1016/j.jbiomech.2010.04.001 
146 
 
Bruijn, S. M., Meijer, O. G., van Dieen, J. H., Kingma, I., & Lamoth, C. J. (2008). Coordination 
of leg swing, thorax rotations, and pelvis rotations during gait: the organisation of total 
body angular momentum. Gait Posture, 27(3), 455-462. doi: 
10.1016/j.gaitpost.2007.05.017 
 
Bullock-Saxton, J. E., Janda, V., & Bullock, M. I. (1994). The influence of ankle sprain injury on 
muscle activation during hip extension. [Clinical Trial Controlled Clinical Trial]. Int J 
Sports Med, 15(6), 330-334. doi: 10.1055/s-2007-1021069 
 
Burton, A. K., Balague, F., Cardon, G., Eriksen, H. R., Henrotin, Y., Lahad, A., . . . van der 
Beek, A. J. (2005). How to prevent low back pain. [Review]. Best Pract Res Clin 
Rheumatol, 19(4), 541-555. doi: 10.1016/j.berh.2005.03.001 
 
Chen, H., Nigg, B. M., Hulliger, M., & de Koning, J. (1995). Influence of sensory input on 
plantar pressure distribution. Clinical Biomechanics, 10(5), 271-274. doi: 
http://dx.doi.org/10.1016/0268-0033(95)99806-D 
 
Chiu, M. C., & Wang, M. J. (2007). Professional footwear evaluation for clinical nurses. Appl 
Ergon, 38(2), 133-141. doi: 10.1016/j.apergo.2006.03.012 
 
Cikajlo, I., & Matjacic, Z. (2007). The influence of boot stiffness on gait kinematics and kinetics 
during stance phase. [Research Support, Non-U.S. Gov't]. Ergonomics, 50(12), 2171-
2182. doi: 10.1080/00140130701582104 
 
Clarke, T. E., Frederick, E. C., & Cooper, L. B. (1983). Effects of shoe cushioning upon ground 
reaction forces in running. Int J Sports Med, 4(4), 247-251. doi: 10.1055/s-2008-1026043 
 
Croce, U. D., Riley, P. O., Lelas, J. L., & Kerrigan, D. C. (2001). A refined view of the 
determinants of gait. Gait Posture, 14(2), 79-84. doi: http://dx.doi.org/10.1016/S0966-
6362(01)00128-X 
 
Crosbie, J., & Vachalathiti, R. (1997). Synchrony of pelvic and hip joint motion during walking. 
Gait &amp; Posture, 6(3), 237-248. doi: 10.1016/s0966-6362(97)00019-2 
 
Crosbie, J., Vachalathiti, R., & Smith, R. (1997a). Age, gender and speed effects on spinal 
kinematics during walking. Gait Posture, 5(1), 13-20. doi: 
http://dx.doi.org/10.1016/S0966-6362(96)01068-5 
 
Crosbie, J., Vachalathiti, R., & Smith, R. (1997b). Patterns of spinal motion during walking. Gait 
&amp; Posture, 5(1), 6-12. doi: 10.1016/s0966-6362(96)01066-1 
 
D'Ambrogi, E., Giacomozzi, C., Macellari, V., & Uccioli, L. (2005). Abnormal foot function in 
diabetic patients: the altered onset of Windlass mechanism. Diabet Med, 22(12), 1713-
1719. doi: 10.1111/j.1464-5491.2005.01699.x 
147 
 
Dingwell, J. B., Cusumano, J. P., Cavanagh, P. R., & Sternad, D. (2001). Local dynamic stability 
versus kinematic variability of continuous overground and treadmill walking. [Clinical 
Trial Research Support, Non-U.S. Gov't]. J Biomech Eng, 123(1), 27-32.  
 
Divert, C., Mornieux, G., Baur, H., Mayer, F., & Belli, A. (2005). Mechanical comparison of 
barefoot and shod running. [Clinical TrialComparative Study]. Int J Sports Med, 26(7), 
593-598. doi: 10.1055/s-2004-821327 
 
Divert, C., Mornieux, G., Freychat, P., Baly, L., Mayer, F., & Belli, A. (2008). Barefoot-shod 
running differences: shoe or mass effect? [Comparative Study]. Int J Sports Med, 29(6), 
512-518. doi: 10.1055/s-2007-989233 
 
Ducroquet, R., Ducroquet, J., & Ducroquet, P. (1968). Walking and limping; a study of normal 
and pathological walking. Philadelphia: Lippincott. 
 
Esenyel, M., Walsh, K., Walden, J. G., & Gitter, A. (2003). Kinetics of high-heeled gait. J Am 
Podiatr Med Assoc, 93(1), 27-32.  
 
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: a flexible statistical 
power analysis program for the social, behavioral, and biomedical sciences. [Research 
Support, Non-U.S. Gov't]. Behav Res Methods, 39(2), 175-191.  
 
Gallagher, S., Marras, W. S., Litsky, A. S., & Burr, D. (2005). Torso flexion loads and the 
fatigue failure of human lumbosacral motion segments. [Research Support, U.S. Gov't, 
Non-P.H.S.]. Spine (Phila Pa 1976), 30(20), 2265-2273.  
 
Gard, S. A., & Childress, D. S. (1997). The effect of pelvic list on the vertical displacement of 
the trunk during normal walking. Gait Posture, 5(3), 233-238. doi: 
http://dx.doi.org/10.1016/S0966-6362(96)01089-2 
 
Gefen, A., Megido-Ravid, M., Itzchak, Y., & Arcan, M. (2002). Analysis of muscular fatigue 
and foot stability during high-heeled gait. Gait Posture, 15(1), 56-63.  
 
Gross, M. L., Davlin, L. B., & Evanski, P. M. (1991). Effectiveness of orthotic shoe inserts in the 
long-distance runner. Am J Sports Med, 19(4), 409-412.  
 
Hamill, J., Russell, E. M., Gruber, A. H., & Miller, R. (2011). Impact characteristics in shod and 
barefoot running. Footwear Science, 3(1), 33-40. doi: 10.1080/19424280.2010.542187 
 
Haywood, K., & Getchell, N. (2009). Life span motor development (5th ed.). Champaign, IL: 
Human Kinetics. 
 
Hodges, P. W., & Richardson, C. A. (1997). Contraction of the abdominal muscles associated 
with movement of the lower limb. [Research Support, Non-U.S. Gov't]. Phys Ther, 77(2), 
132-142; discussion 142-134.  
148 
 
Huang, Y., Meijer, O. G., Lin, J., Bruijn, S. M., Wu, W., Lin, X., . . . van Dieen, J. H. (2010). 
The effects of stride length and stride frequency on trunk coordination in human walking. 
[Research Support, Non-U.S. Gov't]. Gait Posture, 31(4), 444-449. doi: 
10.1016/j.gaitpost.2010.01.019 
 
Huang, Y. P., Bruijn, S. M., Lin, J. H., Meijer, O. G., Wu, W. H., Abbasi-Bafghi, H., . . . van 
Dieen, J. H. (2011). Gait adaptations in low back pain patients with lumbar disc 
herniation: trunk coordination and arm swing. [Research Support, Non-U.S. Gov't]. Eur 
Spine J, 20(3), 491-499. doi: 10.1007/s00586-010-1639-8 
 
Jackson, M., Solomonow, M., Zhou, B., Baratta, R. V., & Harris, M. (2001). Multifidus EMG 
and tension-relaxation recovery after prolonged static lumbar flexion. [Research Support, 
U.S. Gov't, P.H.S.]. Spine (Phila Pa 1976), 26(7), 715-723.  
 
Jackson, N. D., Gutierrez, G. M., & Kaminski, T. (2009). The effect of fatigue and habituation 
on the stretch reflex of the ankle musculature. [Research Support, Non-U.S. Gov't]. J 
Electromyogr Kinesiol, 19(1), 75-84. doi: 10.1016/j.jelekin.2007.06.016 
 
Keenan, G. S., Franz, J. R., Dicharry, J., Della Croce, U., & Kerrigan, D. C. (2011). Lower limb 
joint kinetics in walking: the role of industry recommended footwear. [Comparative 
Study]. Gait Posture, 33(3), 350-355. doi: 10.1016/j.gaitpost.2010.09.019 
 
Kelleher, K. J., Spence, W. D., Solomonidis, S., & Apatsidis, D. (2010). The effect of textured 
insoles on gait patterns of people with multiple sclerosis. Gait Posture, 32(1), 67-71. doi: 
http://dx.doi.org/10.1016/j.gaitpost.2010.03.008 
 
Kendall, F. P. (2005). Muscles : testing and function with posture and pain (5th ed.). Baltimore, 
MD: Lippincott Williams & Wilkins. 
 
Kerrigan, D. C., Todd, M. K., & Riley, P. O. (1998). Knee osteoarthritis and high-heeled shoes. 
[Clinical TrialRandomized Controlled TrialResearch Support, Non-U.S. Gov't]. Lancet, 
351(9113), 1399-1401. doi: 10.1016/S0140-6736(97)11281-8 
 
Kersting, U. G., Janshen, L., Bohm, H., Morey-Klapsing, G. M., & Bruggemann, G. P. (2005). 
Modulation of mechanical and muscular load by footwear during catering. [Research 
Support, Non-U.S. Gov't]. Ergonomics, 48(4), 380-398. doi: 
10.1080/00140130512331332882 
 
Kibler, W. B. (1998). The role of the scapula in athletic shoulder function. [Review]. Am J 
Sports Med, 26(2), 325-337.  
 
Kirtley, C. (2008). History of the study of locomotion: The modern era.  Retrieved January 1, 
2013, from http://www.clinicalgaitanalysis.com/history/modern.html 
 
 
149 
 
Kraus, J. F., Schaffer, K. B., McArthur, D. L., & Peek-Asa, C. (1997). Epidemiology of acute 
low back injury in employees of a large home improvement retail company. [Research 
Support, Non-U.S. Gov't Research Support, U.S. Gov't, P.H.S.]. Am J Epidemiol, 146(8), 
637-645.  
 
Krebs, D. E., Wong, D., Jevsevar, D., Riley, P. O., & Hodge, W. A. (1992). Trunk kinematics 
during locomotor activities. [Research Support, U.S. Gov't, Non-P.H.S.]. Phys Ther, 
72(7), 505-514.  
 
Kristen, K. H., Kastner, J., Holzreiter, S., Wagner, P., & Engel, A. (1998). Funktionelle 
Beurteilung von Kinderschuhen anhand der Ganganalyse von Kindern im Lauflernalter. 
[Biomechanics of Children Shoes using Gait Analyses in Saddlers]. Z Orthop Unfall, 
136(05), 457-462. doi: 10.1055/s-2008-1053684 
 
Kurz, M. J., & Stergiou, N. (2002). Effect of normalization and phase angle calculations on 
continuous relative phase. [Comparative StudyResearch Support, Non-U.S. Gov't]. J 
Biomech, 35(3), 369-374.  
 
Kurz, M. J., & Stergiou, N. (2003). The spanning set indicates that variability during the stance 
period of running is affected by footwear. [Comparative Study]. Gait Posture, 17(2), 132-
135.  
 
Lamoth, C. J., Beek, P. J., & Meijer, O. G. (2002). Pelvis-thorax coordination in the transverse 
plane during gait. [Research Support, Non-U.S. Gov't]. Gait Posture, 16(2), 101-114.  
 
Lamoth, C. J., Meijer, O. G., Wuisman, P. I., van Dieen, J. H., Levin, M. F., & Beek, P. J. 
(2002). Pelvis-thorax coordination in the transverse plane during walking in persons with 
nonspecific low back pain. [Clinical TrialControlled Clinical TrialResearch Support, 
Non-U.S. Gov't]. Spine (Phila Pa 1976), 27(4), E92-99.  
 
Lamoth, C. J. C., Daffertshofer, A., Meijer, O. G., & Beek, P. J. (2006). How do persons with 
chronic low back pain speed up and slow down?: Trunk?pelvis coordination and lumbar 
erector spinae activity during gait. Gait Posture, 23(2), 230-239.  
 
Larsen, K., Weidich, F., & Leboeuf-Yde, C. (2002). Can custom-made biomechanic shoe 
orthoses prevent problems in the back and lower extremities? A randomized, controlled 
intervention trial of 146 military conscripts. [Clinical TrialRandomized Controlled Trial]. 
J Manipulative Physiol Ther, 25(5), 326-331.  
 
Lee, C.-M., Jeong, E.-H., & Freivalds, A. (2001). Biomechanical effects of wearing high-heeled 
shoes. International Journal of Industrial Ergonomics, 28(6), 321-326. doi: 
10.1016/s0169-8141(01)00038-5 
 
Lee, S. J., & Hidler, J. (2008). Biomechanics of overground vs. treadmill walking in healthy 
individuals. Journal of Applied Physiology, 104(3), 747-755. doi: 
10.1152/japplphysiol.01380.2006 
150 
 
Li, J. X., & Hong, Y. (2007). Kinematic and electromyographic analysis of the trunk and lower 
limbs during walking in negative-heeled shoes. [Research Support, Non-U.S. Gov't]. J 
Am Podiatr Med Assoc, 97(6), 447-456.  
 
Lieberman, D. E., Venkadesan, M., Werbel, W. A., Daoud, A. I., D'Andrea, S., Davis, I. S., . . . 
Pitsiladis, Y. (2010). Foot strike patterns and collision forces in habitually barefoot 
versus shod runners. [Research Support, Non-U.S. Gov'tResearch Support, U.S. Gov't, 
Non-P.H.S.]. Nature, 463(7280), 531-535. doi: 10.1038/nature08723 
 
Magee, D. J. (2008). Orthopedic physical assessment (5th ed.). St. Louis, Mo.: Saunders 
Elsevier. 
 
Majumdar, D., Banerjee, P. K., Pal, M., Kumar, R., & Selvamurthy, W. (2006). Temporal spatial 
parameters of gait with barefoot, bathroom slippers and military boots. Indian J Physiol 
Pharmacol, 50(1), 33-40.  
 
Maki, B. E., Perry, S. D., Norrie, R. G., & McIlroy, W. E. (1999). Effect of facilitation of 
sensation from plantar foot-surface boundaries on postural stabilization in young and 
older adults. [Research Support, Non-U.S. Gov'tResearch Support, U.S. Gov't, P.H.S.]. J 
Gerontol A Biol Sci Med Sci, 54(6), M281-287.  
 
Manor, B., & Li, L. (2009). Characteristics of functional gait among people with and without 
peripheral neuropathy. [Research Support, Non-U.S. Gov't]. Gait Posture, 30(2), 253-
256. doi: 10.1016/j.gaitpost.2009.04.011 
 
Margaria, R. (1976). Biomechanics and energetics of muscular exercise. Oxford Eng.: Clarendon 
Press. 
 
Marras, W. S. (2000). Occupational low back disorder causation and control. [Review]. 
Ergonomics, 43(7), 880-902. doi: 10.1080/001401300409080 
 
Marras, W. S., Lavender, S. A., Leurgans, S. E., Fathallah, F. A., Ferguson, S. A., Allread, W. 
G., & Rajulu, S. L. (1995). Biomechanical risk factors for occupationally related low 
back disorders. [Research Support, Non-U.S. Gov'tResearch Support, U.S. Gov't, P.H.S.]. 
Ergonomics, 38(2), 377-410. doi: 10.1080/00140139508925111 
 
McDougall, C. (2009). Born to run : a hidden tribe, superathletes, and the greatest race the 
world has never seen (1st ed.). New York: Alfred A. Knopf. 
 
McGill, S. (2007). Low back disorders : evidence-based prevention and rehabilitation (2nd ed.). 
Champaign, IL: Human Kinetics. 
 
McMullen, J., & Uhl, T. L. (2000). A kinetic chain approach for shoulder rehabilitation. J Athl 
Train, 35(3), 329-337.  
151 
 
Morio, C., Lake, M. J., Gueguen, N., Rao, G., & Baly, L. (2009). The influence of footwear on 
foot motion during walking and running. J Biomech, 42(13), 2081-2088. doi: 
10.1016/j.jbiomech.2009.06.015 
 
Moseley, G. L., Hodges, P. W., & Gandevia, S. C. (2002). Deep and superficial fibers of the 
lumbar multifidus muscle are differentially active during voluntary arm movements. 
[Research Support, Non-U.S. Gov't]. Spine (Phila Pa 1976), 27(2), E29-36.  
 
Mundermann, A., Stefanyshyn, D. J., & Nigg, B. M. (2001). Relationship between footwear 
comfort of shoe inserts and anthropometric and sensory factors. [Clinical Trial 
Randomized Controlled Trial Research Support, Non-U.S. Gov't]. Med Sci Sports Exerc, 
33(11), 1939-1945.  
 
Murley, G. S., Landorf, K. B., Menz, H. B., & Bird, A. R. (2009). Effect of foot posture, foot 
orthoses and footwear on lower limb muscle activity during walking and running: a 
systematic review. [Research Support, Non-U.S. Gov't Review]. Gait Posture, 29(2), 
172-187. doi: 10.1016/j.gaitpost.2008.08.015 
 
Murray, M. P., Drought, A. B., & Kory, R. C. (1964). Walking Patterns of Normal Men. J Bone 
Joint Surg Am, 46, 335-360.  
 
Nashner, L. M. (1980). Balance adjustments of humans perturbed while walking. [Research 
Support, U.S. Gov't, P.H.S.]. J Neurophysiol, 44(4), 650-664.  
 
Nawoczenski, D. A., & Ludewig, P. M. (1999). Electromyographic effects of foot orthotics on 
selected lower extremity muscles during running. [Clinical Trial Randomized Controlled 
Trial Research Support, Non-U.S. Gov't]. Arch Phys Med Rehabil, 80(5), 540-544.  
 
Nurse, M. A., Hulliger, M., Wakeling, J. M., Nigg, B. M., & Stefanyshyn, D. J. (2005). 
Changing the texture of footwear can alter gait patterns. [Clinical Trial Research Support, 
Non-U.S. Gov't]. J Electromyogr Kinesiol, 15(5), 496-506. doi: 
10.1016/j.jelekin.2004.12.003 
 
Nurse, M. A., & Nigg, B. M. (2001). The effect of changes in foot sensation on plantar pressure 
and muscle activity. [Research Support, Non-U.S. Gov't]. Clin Biomech (Bristol, Avon), 
16(9), 719-727.  
 
Oliver, G. D., Dwelly, P. M., Sarantis, N. D., Helmer, R. A., & Bonacci, J. A. (2010). Muscle 
activation of different core exercises. [Research Support, Non-U.S. Gov't]. J Strength 
Cond Res, 24(11), 3069-3074. doi: 10.1519/JSC.0b013e3181d321da 
 
Oliver, G. D., Stone, A. J., & Plummer, H. (2010). Electromyographic examination of selected 
muscle activation during isometric core exercises. [Comparative Study]. Clin J Sport 
Med, 20(6), 452-457. doi: 10.1097/JSM.0b013e3181f7b0ef 
 
152 
 
Opila-Correia, K. A. (1990). Kinematics of high-heeled gait. [Comparative Study]. Arch Phys 
Med Rehabil, 71(5), 304-309.  
 
Parvataneni, K., Ploeg, L., Olney, S. J., & Brouwer, B. (2009). Kinematic, kinetic and metabolic 
parameters of treadmill versus overground walking in healthy older adults. Clinical 
Biomechanics, 24(1), 95-100. doi: http://dx.doi.org/10.1016/j.clinbiomech.2008.07.002 
 
Perry, J. (1992). Gait analysis : normal and pathological function. Thorofare, NJ: SLACK. 
 
Robbins, S. E., & Gouw, G. J. (1990). Athletic footwear and chronic overloading. A brief 
review. [Review]. Sports Med, 9(2), 76-85.  
 
Romer, B., Fox, J., Patel, J., Rehm, J., Shroyer, J., & Weimar, W. (2012). Influence of a Fixed 
Cadence on Shod and Barefoot Gait Kinematics. Paper presented at the 17th Annual Gait 
& Clinical Movement Analysis Society Annual Meeting, Grand Rapids, MI.  
 
Sato, K., Fortenbaugh, D., & Hydock, D. S. (2012). Kinematic changes using weightlifting shoes 
on barbell back squat. J Strength Cond Res, 26(1), 28-33. doi: 
10.1519/JSC.0b013e318218dd64 
 
Saunders, J. B., Inman, V. T., & Eberhart, H. D. (1953). The major determinants in normal and 
pathological gait. J Bone Joint Surg Am, 35-A(3), 543-558.  
 
Savelberg, H. H., Ilgin, D., Angin, S., Willems, P. J., Schaper, N. C., & Meijer, K. (2010). 
Prolonged activity of knee extensors and dorsal flexors is associated with adaptations in 
gait in diabetes and diabetic polyneuropathy. Clin Biomech (Bristol, Avon), 25(5), 468-
475. doi: 10.1016/j.clinbiomech.2010.02.005 
 
Scott, L. A., Murley, G. S., & Wickham, J. B. (2012). The influence of footwear on the 
electromyographic activity of selected lower limb muscles during walking. Journal of 
Electromyography and Kinesiology, 22(6), 1010-1016. doi: 
http://dx.doi.org/10.1016/j.jelekin.2012.06.008 
 
Shakoor, N., Lidtke, R. H., Sengupta, M., Fogg, L. F., & Block, J. A. (2008). Effects of 
specialized footwear on joint loads in osteoarthritis of the knee. [Clinical Trial 
Comparative Study Research Support, N.I.H., Extramural Research Support, Non-U.S. 
Gov't]. Arthritis Rheum, 59(9), 1214-1220. doi: 10.1002/art.24017 
 
Shorten, M., & Mientjes, M. I. V. (2011). The ?heel impact? force peak during running is neither 
?heel? nor ?impact? and does not quantify shoe cushioning effects. Footwear Science, 
3(1), 41-58. doi: 10.1080/19424280.2010.542186 
 
Shroyer, J. F., & Weimar, W. H. (2010). Comparative Analysis of Human Gait While Wearing 
Thong-Style Flip-flops versus Sneakers. J Am Podiatr Med Assoc, 100(4), 251-257.  
 
Smidt, G. L. (1990). Gait in rehabilitation. New York: Churchill Livingstone. 
153 
 
Stergiou, N. (2004). Innovative analyses of human movement. Champaign, IL: Human Kinetics. 
 
Stokes, V. P., Andersson, C., & Forssberg, H. (1989). Rotational and translational movement 
features of the pelvis and thorax during adult human locomotion. [Research Support, 
Non-U.S. Gov't]. J Biomech, 22(1), 43-50.  
 
Taylor, N. F., Evans, O. M., & Goldie, P. A. (2003). The effect of walking faster on people with 
acute low back pain. Eur Spine J, 12(2), 166-172. doi: 10.1007/s00586-002-0498-3 
 
Taylor, N. F., Goldie, P. A., & Evans, O. M. (1999). Angular movements of the pelvis and 
lumbar spine during self-selected and slow walking speeds. Gait Posture, 9(2), 88-94. 
doi: http://dx.doi.org/10.1016/S0966-6362(99)00004-1 
 
Thaut, M. H., Miltner, R., Lange, H. W., Hurt, C. P., & Hoemberg, V. (1999). Velocity 
modulation and rhythmic synchronization of gait in Huntington's disease. Movement 
Disorders, 14(5), 808-819. doi: 10.1002/1531-8257(199909)14:5<808::aid-
mds1014>3.0.co;2-j 
 
van Emmerik, R. E. A., & Wagenaar, R. C. (1996). Effects of walking velocity on relative phase 
dynamics in the trunk in human walking. J Biomech, 29(9), 1175-1184. doi: 
http://dx.doi.org/10.1016/0021-9290(95)00128-X 
 
van Gent, R. N., Siem, D., van Middelkoop, M., van Os, A. G., Bierma-Zeinstra, S. M., & Koes, 
B. W. (2007). Incidence and determinants of lower extremity running injuries in long 
distance runners: a systematic review. [Review]. Br J Sports Med, 41(8), 469-480; 
discussion 480. doi: 10.1136/bjsm.2006.033548 
 
Wagenaar, R. C., & Beek, W. J. (1992). Hemiplegic gait: A kinematic analysis using walking 
speed as a basis. J Biomech, 25(9), 1007-1015. doi: http://dx.doi.org/10.1016/0021-
9290(92)90036-Z 
 
Wagenaar, R. C., & van Emmerik, R. E. A. (2000). Resonant frequencies of arms and legs 
identify different walking patterns. J Biomech, 33(7), 853-861. doi: 
http://dx.doi.org/10.1016/S0021-9290(00)00020-8 
 
Wakeling, J. M., & Liphardt, A. M. (2006). Task-specific recruitment of motor units for 
vibration damping. [Research Support, Non-U.S. Gov't]. J Biomech, 39(7), 1342-1346. 
doi: 10.1016/j.jbiomech.2005.03.009 
 
Wakeling, J. M., Liphardt, A. M., & Nigg, B. M. (2003). Muscle activity reduces soft-tissue 
resonance at heel-strike during walking. [Clinical Trial Comparative Study Randomized 
Controlled Trial Research Support, Non-U.S. Gov't]. J Biomech, 36(12), 1761-1769.  
 
Wegener, C., Hunt, A. E., Vanwanseele, B., Burns, J., & Smith, R. M. (2011). Effect of 
children's shoes on gait: a systematic review and meta-analysis. J Foot Ankle Res, 4, 3. 
doi: 10.1186/1757-1146-4-3 
154 
 
Whittle, M. (2003). Gait analysis : an introduction (3rd ed.). Edinburgh ; New York: 
Butterworth-Heinemann. 
 
Willson, J. D., Dougherty, C. P., Ireland, M. L., & Davis, I. M. (2005). Core stability and its 
relationship to lower extremity function and injury. [Review]. J Am Acad Orthop Surg, 
13(5), 316-325.  
 
Wu, W., Meijer, O. G., Jutte, P. C., Uegaki, K., Lamoth, C. J., Sander de Wolf, G., . . . Beek, P. 
J. (2002). Gait in patients with pregnancy-related pain in the pelvis: an emphasis on the 
coordination of transverse pelvic and thoracic rotations. [Clinical Trial Comparative 
Study Controlled Clinical Trial Research Support, Non-U.S. Gov't]. Clin Biomech 
(Bristol, Avon), 17(9-10), 678-686.  
 
Wu, W., Meijer, O. G., Lamoth, C. J., Uegaki, K., van Dieen, J. H., Wuisman, P. I., . . . Beek, P. 
J. (2004). Gait coordination in pregnancy: transverse pelvic and thoracic rotations and 
their relative phase. [Clinical Trial Controlled Clinical Trial Research Support, Non-U.S. 
Gov't]. Clin Biomech (Bristol, Avon), 19(5), 480-488. doi: 
10.1016/j.clinbiomech.2004.02.003 
 
Wu, W. H., Meijer, O. G., Bruijn, S. M., Hu, H., van Dieen, J. H., Lamoth, C. J., . . . Beek, P. J. 
(2008). Gait in Pregnancy-related Pelvic girdle Pain: amplitudes, timing, and 
coordination of horizontal trunk rotations. [Research Support, Non-U.S. Gov't]. Eur 
Spine J, 17(9), 1160-1169. doi: 10.1007/s00586-008-0703-0 
 
Wyke, B. (1967). The neurology of joints. Ann R Coll Surg Engl, 41(1), 25-50.  
 
Xu, Y., Bach, E., & Orhede, E. (1997). Work environment and low back pain: the influence of 
occupational activities. Occup Environ Med, 54(10), 741-745.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
155 
 
Appendix A 
Health Screening Questionnaire 
Please read each question carefully and answer honestly. If you do not understand the 
question, please ask the investigator for clarification. Check the appropriate answer.  
Participant Number: _____________ 
Yes  No 
____  ____ 1) Are you under the age of 19? 
____  ____ 2) Are you over the age of 35? 
____  ____ 3) Do you currently have an injury that prevents or hinders your ability to  
walk normally in shoes? 
____  ____ 4) Do you currently have an injury that prevents or hinders your ability to  
walk normally when barefoot? 
____  ____ 5) Have you had any injuries or surgeries to your lower extremity, pelvis, 
low back, or trunk during the past year? 
____  ____ 6) Do you currently have any pain in your lower extremity, pelvis, 
low back, or trunk? 
____  ____  7) Do you have any reason to believe that your participation in this  
investigation may put your health or well-being at risk? 
                         8) Have you ever been told you have an inner ear disorder? 
_____   _____   9) Has your doctor ever said that you have heart trouble?  
                          10) Are you currently taking any medication that you think might  
influence your ability to participate in this study? 
                           11) Are you allergic to adhesives? 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
156 
 
Appendix B 
 
(NOTE:  DO NOT SIGN THIS DOCUMENT UNLESS AN IRB APPROVAL STAMP WITH 
CURRENT DATES HAS BEEN APPLIED TO THIS DOCUMENT.) 
INFORMED CONSENT 
for a Research Study entitled 
?Influence of Tactile Feedback on Trunk Coordination and  
Electromyography During Walking? 
 
You are invited to participate in a research study to investigate the influence of tactile 
feedback on trunk coordination and electromyography during walking. The study is being 
conducted by Braden Romer under the supervision of Dr. Wendi Weimar, Associate Professor, 
in the Auburn University Department of Kinesiology. You were selected as a potential 
participant because you are between the ages 19 and 35, and your health condition might, 
through a pre-screening health questionnaire to follow, permit you to perform the conditions 
safely and successfully. The purpose of this study is to gain a greater scientific understanding of 
the variables responsible for alterations in trunk coordination. Furthermore, the effect of variable 
sensory inputs to the foot on changes in walking patterns will be investigated.  
 
What will be involved if you participate? If you decide to participate in this research study, 
you will be asked to come to the Sports Biomechanics Laboratory (Memorial Coliseum 1127) for 
two testing sessions, lasting approximately 60 minutes each. Paperwork will only need to be 
filled out during the initial session. After meeting participation criteria, your mass, height, leg 
length and arch height will be measured for normalization purposes. Next you will be fitted for 
the proper shoe and insole size. During both sessions, you will then be asked to change into 
spandex shorts and spandex tank tops to minimize any potential interference from personal 
clothing. Researchers will need to identify six separate muscles (right erector spinae, right and 
left gluteus medius, right biceps femoris, right soleus, and right peroneus longus). Identification 
of the pertinent muscles will occur through a combination of palpation and manual muscle 
testing, specific to each muscle: 
 
? The right erector spinae will be identified during a prone (face down), lumbar extension 
(bending backwards) task. 
? The right gluteus medius will be identified with you lying on your left side while 
concurrently attempting a hip abduction, similar to a jumping jack, task.  
? The left gluteus medius will be identified with you lying on your right side while 
concurrently attempting a hip abduction, similar to a jumping jack, task. 
? The right biceps femoris will be identified during a prone (face down), hip extension 
(pushing your leg backwards) task. 
 
Participant?s Initials__________
157 
 
 
? The soleus will be identified during a seated, plantarflexion (pushing toes down) task 
with your knees bent to 90?. 
? The peroneus longus will be identified during a task involving seated plantarflexion and 
eversion (pushing toes down while flattening your foot).  
 
Once all muscles have been appropriately identified, the skin area around the electrode 
placement site will be shaved, abraded, and cleaned with a 70% alcohol solution to reduce the 
likelihood of electrical impedance. Electrodes will be placed near the center of the muscle belly, 
and parallel to the direction of the muscle fibers for the muscles of interest. A ground electrode 
will be placed on the body aspect of the tibial tuberosity (upper shin). A thin, nylon type wrap 
will be placed over the electrodes and limb segment to assist in preventing electrode movement. 
A maximum voluntary isometric (MVIC) will then be performed and used for normalization 
purposes. 
 
? An MVIC for the erector spinae will be recorded during a prone lumbar extension 
task (bending backwards).  
? An MVIC for the right gluteus medius will be recorded during a left side-laying, hip 
abduction task (similar to a jumping jack). 
? An MVIC for the left gluteus medius will be recorded during a right side-laying, hip 
abduction task (similar to a jumping jack). 
? An MVIC for the biceps femoris will be recorded during supine (face up), hip 
extension task (pushing leg backwards).  
? An MVIC for the soleus and peroneus longus will be recorded simultaneously during 
a seated, plantarflexion (pushing toes down) task with knees bent to 90?. 
 
The final step of preparation will be to affix 37 markers to the skin and clothing of the 
participant with double-sided tape. Marker placement will be in accordance with the Vicon Plug-
In-Gait model. Next, a static capture will be performed with you standing in the center of the 
capture volume, upright, and arms out to the side. During the initial visit, ordering of conditions 
will be randomized based on cadence (the number of steps per minute) or velocity (how fast you 
walk). During the first testing session (day), only one variable will be manipulated, while during 
the second testing session (day) the other factor will be manipulated. Next, footwear order will 
be randomized (barefoot, shoe, textured insole, shoe with textured insole). After a five minute 
footwear acclimation period, you will be asked to perform three trials of your preferred walking 
style through the center of the capture volume. Next, you will be asked to perform three 
additional trials at either 125% or 75% of free-walking cadence or velocity. After those three 
trials have been successfully performed, you will be asked to complete three additional trials at 
the other parameter (125% or 75%). A rest period of a minimum of five minutes will be provided 
to you after each the three conditions. After the rest period, your footwear will be changed and 
you will be provided a minimum of five minutes in which to acclimate to the specific footwear 
condition. The same data collection procedures (free-walking, 125% and 75% of velocity or 
cadence) will take place until successful completion for each of the four footwear conditions. 
The second day of testing will follow the exact same protocol as the first day, except there will 
not be a time period allocated to the completion of paperwork.   
Participant?s Initials__________ 
158 
 
 
Are there any risks or discomfort? It is possible that you may sustain muscle soreness or a 
muscle injury. However, the risks associated with this study should not be any more than 
normally encountered in normal activities of daily living. Other risks include a possible adverse 
response to the adhesive on the electrodes and reflective markers; however, these risks are 
similar to risks associated with applying a Band-Aid.  In addition, these risks should be 
minimized through the use of a health screening questionnaire.  Last, the MVIC could result in a 
musculoskeletal injury such as a strain, sprain or muscle soreness, which is possible in any type 
of lifting or athletic activity. In the unlikely event that you sustain an injury from participation in 
this study, you will be required to assume full financial responsibility for your own medical care. 
Participants are responsible for any and all medical cost resulting from injury during the study.  
 
Are there any benefits to yourself or others? There is no direct benefit to you; however, you 
may request a copy of the written results upon the completion of the study. 
 
If you change your mind about participating, you can withdraw at any time during the study. 
Your participation is completely voluntary. If you choose to withdraw, your data can be 
withdrawn as long as it is identifiable. Your decision about whether or not to participate or stop 
participating will not jeopardize your future relations with Auburn University, Department of 
Kinesiology, or the Sport Biomechanics Laboratory.  
 
You privacy will be protected. Any information obtained in connection with this study will 
remain confidential as the data will correspond solely to an identification number. Information 
obtained through your participation may be published in a scientific journal and/ or presented at 
a scholarly conference.  
 
If you have questions about this study, please ask them now or contact Braden Romer at 
bhr0002@auburn.edu, Dr. Wendi Weimar at weimawh@auburn.edu, or by calling the Sport 
Biomechanics Laboratory at 334.844.1468. A copy of this document will be given to you to 
keep. 
 
If you have questions about your rights as a research participant, you may contact the 
Auburn University of Human Subject Research of the Institutional Review Board by phone (334) 
844-5966 or email at hsubjec@auburn.edu or IRBChair@auburn.edu. 
 
HAVING READ THE INFORMATION PROVIDED, YOU MUST DECIDE WHETHER 
OR NOT YOU WISH TO PARTICIPATE IN THIS RESEARCH STUDY. YOUR 
SIGNATURE INDICATED YOUR WILLINGNESS TO PARTICIPATE. 
 
_____________________________   ____________________________________ 
Printed Name   Date   Investigator obtaining consent            Date 
 
 
_________________________  ____________________________________ 
Participant?s signature       Printed Name