- Email: [email protected]
Biomechanical analysis of gait waveform data: exploring differences between shod and barefoot running in habitually shod runners
Accepted Manuscript Title: Biomechanical analysis of gait waveform data exploring differences between shod and barefoot running in habitually shod runners. Authors: Nicholas Tam, Danielle Prins, Nikhil V. Divekar, Robert P. Lamberts PII: DOI: Reference:
S0966-6362(17)30849-4 http://dx.doi.org/10.1016/j.gaitpost.2017.08.014 GAIPOS 5762
To appear in:
Gait & Posture
Received date: Revised date: Accepted date:
17-3-2017 8-8-2017 11-8-2017
Please cite this article as: Tam Nicholas, Prins Danielle, Divekar Nikhil V, Lamberts Robert P.Biomechanical analysis of gait waveform data exploring differences between shod and barefoot running in habitually shod runners.Gait and Posture http://dx.doi.org/10.1016/j.gaitpost.2017.08.014 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Biomechanical analysis of gait waveform data exploring differences between shod and barefoot running in habitually shod runners.
Nicholas Tama, Danielle Prinsa, Nikhil V. Divekara, Robert P. Lambertsa,b.
a-
Division of Exercise Science and Sports Medicine, Department of Human Biology,
Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa. b
-Institute of Sport and Exercise Medicine, Division of Orthopaedic Surgery,
Department of Surgical Sciences, Faulty of Medicine and Health Sciences, Stellenbosch University, Tygerberg, South Africa.
Word count: 3077 Abstract count: 258 Number of Figures: 3 Number of Tables: 2
Address for correspondence: Nicholas Tam, Division of Exercise Science and Sports Medicine, Department of Human Biology Faculty of Health Sciences, University of Cape Town 3rd Floor, The Sports Science Institute of South Africa, Boundary Road, Newlands, 7725 Cape Town, South Africa [email protected]
Highlights
Differences between barefoot and shod gait also occur in swing phase. Lower limb joint co-ordination appears to differ between barefoot and shod gait Habitual shod runners exhibit earlier movement initiation when barefoot than shod Neuromuscular control in response to differing footwear requires further research. Statistical parametric mapping may provide important insight into gait where previously ignored.
Abstract: The aim of this study was to utilise one-dimensional statistical parametric mapping to compare differences between biomechanical and electromyographical waveforms in runners when running in barefoot or shod conditions. Fifty habitually shod runners were assessed during overground running at their current 10-km race running speed. Electromyography, kinematics and ground reaction forces were collected during these running trials. Joint kinetics were calculated using inverse dynamics. One-dimensional statistical parametric mapping one sample t-test was conducted to assess differences over an entire gait cycle on the variables of interest when barefoot or shod (p<0.05). Only sagittal plane differences were found between barefoot and shod conditions at the knee during late stance (18-23% of the gait cycle) and swing phase (74-90%); at the ankle early stance (0-6%), mid-stance (28-38%) and swing phase (81-100%). Differences in sagittal plane moments were also found at the ankle during early stance (2, 4-5%) and knee during early stance (5-11%). Condition differences were also found in vertical ground reaction force during early stance between (3-10%). An acute bout of barefoot running in habitual shod runners invokes temporal differences throughout the gait cycle. Specifically, a co-ordinative responses between the knee and ankle joint in the sagittal plane with a delay in the impact transient peak; onset of the knee extension and ankle plantarflexion moment in the shod compared to barefoot condition was found. This appears to affect the delay in knee extension and ankle plantarflexion during late stance. This study provides a glimpse into the co-ordination of the lower limb when running in differing footwear.
Key words: biomechanics; footwear; exercise; injury; gait.
Introduction: The statistical analysis of running gait is often restricted to the definition and extraction of discrete parameters of the gait pattern such as peaks, ranges and instances considered meaningful. Various methods have previously been employed such as principal component analysis and support vector machine, but both are pattern recognition techniques used to extract important data from large datasets [1,2]. These techniques are complex and often require large datasets to identify small changes. Recently, the use of one-dimensional statistical parametric mapping has been proposed to assess biomechanical waveform data considering the temporal nature of these data sets[3,4]. In the case of describing barefoot running gait the variables of interest largely focus on discrete loci, such as the joint kinematics at initial ground contact, which include foot strike pattern and the initial loading rate [5–7]. In addition, data often are expressed and summarized over certain periods of the running gait, for example range of motion or weight transfer during the stance phase. Based on the assumption that risk factors are most strongly correlated to the ground contact phase, most running injury research has focused on parameters measured between initial ground contact and the peak vertical ground reaction force [8].Differences in footstrike pattern at initial ground contact when running in running shoes or barefoot have been of a special interest for many researchers especially with regards to its relationship with initial loading rate [9,10]. However, recent studies have reported conflicting differences in the initial loading rate of the vertical ground reaction forced between barefoot and shod runners [5,7,11]. Although the reasoning for researching biomechanical variables between initial ground contact and peak vertical ground reaction force are meritorious, no study to date has researched if other biomechanical events outside of this period contribute to footwear condition differences. Early or late changes in joint biomechanics can potentially contribute to explaining the differences around the initial landing phase. It would be of interest to describe the periods surrounding the above-mentioned epoch and further determine the instances where differences begin and end on a temporal scale. This potentially will assist researchers in determining the period in the gait cycle that may influence these discrete events of interest. It also would be of interest to describe the differences in muscle activity that may are associated with running in the barefoot or shod condition, as peak and average amplitude differences have been found previously in habitually shod runners when running barefoot or shod [12–
14]. Where lower tibialis anterior, lateral gastrocnemius and biceps femoris activity was found when barefoot over the entire period of stance. The aim of this study is to describe the differences between barefoot and shod kinematics, kinetics and electromyography over the gait cycle in habitually shod runners using one-dimensional statistical parametric mapping (1DSPM) [3]. Specifically, we aim to investigate variables during swing that may discriminate between the two footwear conditions. It was hypothesized that kinematic differences between shod and barefoot running would be found beyond stance phase and that kinetic differences between conditions would be restricted to the this period. Methods: Participants: Fifty habitually shod (traditional cushioned shoes) runners volunteered to participate in this study. Participants were able to run 10km in <50 minutes and were injury free for six months prior to the study. Participants provided written informed consent and were fully aware of the benefits and potential risks associated with the study. The study was granted ethical approval by the Human Research Ethics Committee of the study institution and adheres the principles laid down in the Declaration of Helsinki (2013). Experimental conditions: Biomechanical testing was conducted under two different conditions, namely (1) barefoot running condition and (2) in the running shoe in which they were currently completing the most training mileage. All shod midsoles comprised of traditional ethylene vinyl acetate (EVA) cushioned shoes and were not controlled for mileage, shore count or heel-toe drop except in the case of a marketed minimalist shoe. The runners were afforded a familiarization (two lengths of the running track) in condition before performing the running trials Instrumentation: Running trials were conducted on a 40 m indoor synthetic running track. Threedimensional marker trajectories were captured using an 8-camera VICON MX motion analysis system (Oxford Metrics Ltd, Oxford, UK), sampling at 250 Hz. Ground reaction force (GRF) data were collected using a 900x600 mm force platform (AMTI, Watertown, MA, USA), sampling at 2000 Hz. Sixteen Reflective markers were attached according to a modified Helen-Hayes Marker set and the lower body PlugInGait model was applied.
Surface electromyography was measured for four lower limb muscles, namely rectus femoris (RF), biceps femoris (BF), tibialis anterior (TA) and gastrocnemius medius (MG). Prior to placement, the skin areas were prepared and two surface electrodes placed on the muscle location according to SENIAM guidelines[15] Leads and preamplifiers connected to the electrodes were secured with medical grade tape to avoid artefacts from lower limb movement during running. The transmitter unit was secured in a harness strapped to the participant’s back and data sampled at 2000 Hz (Noraxon 2400T G2, Noraxon, USA). Procedures: Participants completed 6 clean overground running trials in each footwear condition in a randomized order, with no instruction to running style. The speed of overground running trials was set based on the participant’s current 10km performance pace (within a month). Trials were accepted if the velocity was within ± 5% of the target speed. During these runs, synchronised collection of marker motion, force platform and electromyography measurements were obtained, a successful trial was defined as one within the specified velocity range, where all markers were in view of the cameras and there was no visual evidence of force platform targeting. After completing running trials at the designated speeds, the runners then completed three maximal sprints down the 60 m runway in the shod condition. Data analysis: Marker trajectory and force platform data were filtered using a low-pass fourth-order Butterworth filter with a cut-off frequency at 20 and 100 Hz respectively. For each trial, one complete stance phase of the gait cycle was analysed. Three-dimensional lower extremity joint angles and net resultant joint moments using a Newton-Euler inverse dynamics approach were calculated [16,17]. Three-dimensional joint moments were expressed as external moments normalized to body mass (Nmkg-1). Raw digital EMG signal of both sub-maximal and sprint trials were processed as previously described [18], through rectification and root mean squared smoothing window at 50 ms. Subsequently, sub-maximal trials were normalized to peak activity from the appropriate trial. The data for each participant’s right limb were averaged over 6 trials for each condition. Sagittal, and frontal plane ankle and knee angles (degrees) and moments (Nmkg-1) are reported. Specifically, all muscle activity, kinematic and kinetic data are represented as waveforms that changed continuously throughout the gait cycle and were defined with 101 data points, one for each percentage of the cycle.
Statistical analysis Data were screened for normality of distribution using a Shapiro-Wilk’s Normality test. Differences in initial loading rate; ground contact time; knee and ankle joint stiffness were assessed using t-tests or non-parametric Wilcoxon signed rank test. To detect significant differences between the kinematic and kinetic waveforms in an objective way, 1DSPM was employed [3]. The barefoot and shod conditions were compared to each other using an one-way ANOVA (SPM{f}) All 1DSPM analyses were implemented using the open-source 1DSPM code (v.M0.1, www.spm1d.org) in Matlab (R2014a, 8.3.0.532, Mathworks Inc., Natick, MA, USA). Significance was set at p<0.05. Results: The characteristics of the group of runners were age: 30.3 ± 7.4 years, mass: 71.5 ± 11.1 kg, height: 1.8 ± 1.0 m and recent 10-km time: 48.3 ± 7.8 min. The mean running speed during over-ground running trials was 3.5 ± 0.5 ms-1 and no differences in running speed between the footwear conditions were found (3.5 ± 0.4 ms-1 vs. 3.5 ± 0.6 ms-1 for shod and barefoot, respectively). Both vertical initial loading rate and ankle stiffness were greater in the barefoot running condition than in the shod running conditions, while ground contact time was shorter when running barefoot (Table 1). Although no waveform difference were found in the frontal and transfers planes between the barefoot or shod condition, waveform differences were found in the sagittal plane (see Table 2 and Figure 1). Kinematics A notably reduced knee flexion was found between 18-23% of the gait cycle (F=8.581; P=0.038) in the barefoot condition than when running shod. In line with this an earlier knee extension response during late swing (74-90% of the gait cycle) was seen in the barefoot condition (F=8.581; P=0.004). In addition, reduced ankle dorsiflexion was observed in the barefoot running condition during the 0-6% of the gait cycle (F=8.854; P=0.032) (Table 2 and Figure 1). During weight transfer ankle plantarflexion between 28-38% of the gait cycle (F=8.854; P=0.013) and during late swing at 81-100% of the gait cycle (F=8.854; P<0.001) was greater in the barefoot when compared to shod condition. Forces and moments
Differences in the vertical ground reaction force were found at early onset of stance where the rise in force when running barefoot was earlier than shod at 3-11% of the gait cycle (F=8.820; P=0.027) (Table 2 and Figure 2). Greater barefoot knee extension moments were found at 2 and 4-5% of the gait cycle (F=11.737; P=0.47 and P=0.035 respectively). Alongside this, greater barefoot ankle plantarflexion moments were found between 5-11% of the gait cycle (F=11.661; P<0.001). Muscle activity No differences in muscle activity waveforms were found between the two conditions (Figure 3). Discussion: The novelty of this study is that differences between barefoot and shod gait have been found throughout the gait cycle including swing phase using statistical parametric mapping. In addition, the analysis technique was able to identify the temporal periods of the gait cycle. This expands our knowledge of changes in gait associated with different footwear that may potentially impact on the understanding of function, injury and performance, as previous footwear research has largely focused on ground contact phases. Swing Phase Swing phase of gait is often overlooked, as it does not directly assess the interaction between the moving body and it’s surrounding environment [19]. Hence, the ground contact phase of gait has mainly been studied as a consequence of the main epoch where large forces are experienced during gait. However, during terminal swing of the gait cycle (74-90%), barefoot knee extension started earlier and was greater than when running in the shod condition. In addition, it appears that peak knee flexion during swing may have been slightly reduced when running barefoot, however this difference was non-significant. Coinciding with the earlier knee extension response, we found that the ankle during barefoot running remains in a neutral position during terminal swing phase of the gait cycle (81-100%), while the ankle is dorsiflexed during this phase when running in a shod condition. These findings reflect an adaptive barefoot running style, which allows the barefoot runner to have a more mid-foot landing pattern. In the shod condition, more knee flexion and ankle dorsiflexion is needed to allow adequate foot ground clearance and a heel strike landing pattern [20]. The more mid-foot landing pattern in the barefoot running condition is also confirmed by the early phase of the stance phase (0-6% of the gait cycle|) when the ankle remains neutral in the barefoot condition and is still in maximal
dorsiflexion in the shod running condition. These findings support a mid/forefoot landing in the barefoot running condition and a heelstrike footstrike pattern in the shod runnning condition [21,22]. In addition, this footwear difference also illustrates the influence of shoe cushioning/construction on gait as the ankle is required to preemptively dorsiflex prior ground contact in order to clear the ground during swing as a result of the medium (footwear cushioning) between the foot and the ground. Ground Contact Phase The first difference found at ground contact was ankle plantar-dorsiflexion from 0-6% of the gait cycle. The barefoot condition remains in neutral from swing phase and the shod condition in dorsiflexion throughout initial ground contact where both conditions merge at foot flat phase and initial impact peak. Prior to ground contact the barefoot condition remains in a neutral angle during terminal swing and this persists at initial ground contact. This observation at initial stance phase has previously been documented in previous research assessing discrete variables [7,19]. However, contrary to previous research, this study found that temporal differences continue over a quarter of the gait cycle. This difference is a result of runners striving for a mid/forefoot landing when barefoot as this would reduce the discomfort of a rear foot landing [19]. We can postulate that runners alter their foot strike pattern as a result of the cushioning of the shoe giving them the ease to land more rearfoot [19,22]. Concurrently, the barefoot vertical ground reaction force produced a earlier but lower force at 3-10% of the gait cycle, while a later but higher ground reaction force was seen in the shod running condition. Ankle plantarflexion influences the vertical ground reaction force since a mid/forefoot landing creates a greater landing surface [6,23] and consequently the impact is distributed across a larger surface area resulting in a rise in peak. Further, this also informs us of the neuromuscular anticipation and preparation required for landing in the absence of footwear cushioning, despite not finding any muscle activity differences between the two conditions. As previously stated an earlier rise in the vertical ground reaction force was found in the barefoot condition, this was associated with a greater initial loading rate too. A greater initial loading rate has previously been linked to risk factors for tibial stress factors because recent evidence has suggest that the shank and foot mass interaction largely influence the initial ground contact kinetics and up to impact transient peak [23–25]. The impact transient peaks between the two conditions are occur at differing times of the gait cycle and maybe different in magnitude. Thus, caution may still be advised on acute transition to barefoot running as this rate of
force development during early stance may increase injury risk in the shank-foot complex. Differences in ankle sagittal kinematics and vertical ground reaction force intersect and proceed with a greater barefoot knee extension moment and ankle plantarflexion moment between 2; 3-5% and 5-11% of the gait cycle, respectively. A larger barefoot knee extension moment was observed prior to the difference in ankle plantarflexion moment. These findings inform us of the kinematic response to the lower limb landing mechanism and suggest the influence of neuromuscular control in the absence of footwear. This earlier and greater barefoot knee extension moment is an indicator of the eccentric action of the quadriceps in response to absent footwear cushioning and neutral ankle flexion angle at ground contact. Subsequently, the onset of shod ankle plantarflexion moment was delayed as a consequence of differing kinematic landing strategies because the shod ankle enters stance already in dorsiflexion. Therefore, the need to increase the plantarflexion moment is reduced or delayed when shod compared to the barefoot condition. The onset of the barefoot plantarflexion moment typically occurs between 5-10% of the gait cycle [26]. Following this difference in ankle flexion during early stance, greater shod knee flexion between 18-23% of the gait cycle when compared to the barefoot condition. This finding suggests that in the shod condition knee flexion is both greater and maintained through weight transfer of stance. This finding suggests a joint kinematic response to weight acceptance and transfer in the presence or lack of footwear cushioning on the joint mechanics [7]. Interestingly, it appears that this increased knee flexion in the shod condition maybe a response to the greater ankle stiffness as result of the greater dorsiflexion of the ankle at the beginning of stance that reduces ankle range of motion. Subsequently, this may also impact ankle plantarflexion during late stance and towards toe-off as a difference was found between 28-38% of the gait cycle. This difference appears to be a result of delayed plantarflexion in the shod condition, as a consequence of delayed knee extension from peak flexion during weight transfer, as peak knee flexion during stance appears to be sustained and is greater (difference between 18-23% of the gait cycle) than barefoot in the shod condition. This difference also reflects a delayed response in ankle plantarflexion in the shod condition, with the barefoot condition obtaining an earlier peak plantarflexion angle at toe-off prior to toe-off.
Differences in ground contact time were found although all trials were performed at the same speed. Whether or not this may impact on the differences found in this study are unknown, however, the entire gait cycle was normalised to 101 points for ease of interpretation and reference. Further, no differences were found between conditions in muscle activity. Previous researchers have found amplitude differences over the differing phases of gait in barefoot and shod conditions [12,13,22,27]. However, many of these studies focus on differing footfall patterns and imposing differing gait patterns. Thus, the instinctive adaptation is often ignored or not of interest, whereas this study provided no such feedback and segregation of footfall patterns. Ultimately, the high variability of the muscle activity data did not allow this technique to infer whether the data sets were indeed different from random. Further, an important consideration regarding this study is that the commonly used PlugInGait model has been found to reproduce highly variable transverse plane angles. This limits determining clinically relevant changes/differences between footwear conditions [28], thus the use of a different model may reveal differences not found in this study. This study remains relevant as it suggests to researchers to refrain from over simplifying gait data and to appreciate the entire waveform and further supports analyses like 1DSPM to better interpret and understand changes in gait. Conclusion: This study provides insight into terminal swing, ground contact preparation phase and loading rate dynamics when running in a shod or barefoot condition. Differences between barefoot and shod were mainly seen in the knee and ankle joint in the sagittal plane and landing forces. Notable footwear differences were seen during late swing and early landing phase. Barefoot running was characterised by a mid-/forefootstrike pattern, whereby preparation starts during terminal swing phase exhibited by an earlier knee extension response and a more neutral ankle flexion position during the final part of the swing and early stance phase. These subtle differences seem to reflect complex motor control that adapts gait in different footwear conditions. However, as running in different conditions alters gait, runners should take slowly introduce new running styles rather than drastically swop over. Conflicts of interest: No conflict of interest
References: [1]
Trudeau MB, von Tscharner V, Vienneau J, Hoerzer S, Nigg BM. Assessing Footwear Effects from Principal Features of Plantar Loading during Running. Med Sci Sports Exerc 2015;47:1988–96. doi:10.1249/MSS.0000000000000615.
[2]
Astephen JL, Deluzio KJ. A multivariate gait data analysis technique: application to knee osteoarthritis. Proc Inst Mech Eng H 2004;218:271–9.
[3]
Pataky TC, Vanrenterghem J, Robinson MA. Zero- vs. one-dimensional, parametric vs. non-parametric, and confidence interval vs. hypothesis testing procedures in one-dimensional biomechanical trajectory analysis. J Biomech 2015;48:1277–85. doi:10.1016/j.jbiomech.2015.02.051.
[4]
Nedergaard NJ, Robinson MA, Eusterwiemann E, Drust B, Lisboa PJ, Vanrenterghem J. The Relationship Between Whole-Body External Loading and Body-Worn Accelerometry During Team Sports Movements. Int J Sports Physiol Perform 2016:1–23. doi:10.1123/ijspp.2015-0712.
[5]
Willy RW, Davis IS. Kinematic and kinetic comparison of running in standard and minimalist shoes. Med Sci Sports Exerc 2014;46:318–23. doi:10.1249/MSS.0b013e3182a595d2.
[6]
Lieberman DE, Venkadesan M, Werbel WA, Daoud AI, D’Andrea S, Davis IS, et al. Foot strike patterns and collision forces in habitually barefoot versus shod runners. Nature 2010;463:531–5. doi:10.1038/nature08723.
[7]
Tam N, Astephen Wilson JL, Coetzee DR, van Pletsen L, Tucker R. Loading rate increases during barefoot running in habitually shod runners: Individual responses to an unfamiliar condition. Gait Posture 2016;46:47–52. doi:10.1016/j.gaitpost.2016.02.013.
[8]
Tam N, Astephen Wilson JL, Noakes TD, Tucker R. Barefoot running: an evaluation of current hypothesis, future research and clinical applications. Br J Sports Med 2014;48:349–55. doi:10.1136/bjsports-2013-092404.
[9]
Nigg BM. The role of impact forces and foot pronation: a new paradigm. Clin J Sport Med 2001;11:2–9.
[10]
Davis IS, Bowser BJ, Mullineaux DR. Greater vertical impact loading in female runners with medically diagnosed injuries: a prospective investigation. Br J Sports Med 2016;50:887–92. doi:10.1136/bjsports-2015-094579.
[11]
Rice HM, Jamison ST, Davis IS. Footwear Matters: Influence of Footwear and Foot Strike on Loadrates During Running. Med Sci Sports Exerc 2016:2462–8. doi:10.1249/MSS.0000000000001030.
[12]
Ervilha UF, Mochizuki L, Jr AF. Are muscle activation patterns altered during shod and barefoot running with a forefoot footfall pattern ? J Sports Sci 2016;0:1–7. doi:10.1080/02640414.2016.1231410.
[13]
Tam N, Tucker R, Astephen Wilson JL. Individual Responses to a Barefoot Running Program: Insight Into Risk of Injury. Am J Sports Med 2016;44:777– 84. doi:10.1177/0363546515620584.
[14]
Snow NJ, Basset FA, Byrne J. An Acute Bout of Barefoot Running Alters Lower-limb Muscle Activation for Minimalist Shoe Users. Int J Sports Med 2016;37:382–7. doi:10.1055/s-0035-1565140.
[15]
Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 2000;10:361–74.
[16]
Winter DA. Overall principle of lower limb support during stance phase of gait. J Biomech 1980;13:923–7. doi:10.1016/0021-9290(80)90162-1.
[17]
Grood ES, Suntay WJ. A joint coordinate system for the clinical description of three-dimensional motions: application to the knee. J Biomech Eng 1983;105:136–44. doi:6865355.
[18]
Albertus-Kajee Y, Tucker R, Derman W, Lamberts RP, Lambert MI. Alternative methods of normalising EMG during running. J Electromyogr Kinesiol 2011;21:579–86. doi:10.1016/j.jelekin.2011.03.009.
[19]
DeWit B, De Clercq D, Aerts P. Biomechanical analysis of the stance phase during barefoot and shod running. J Biomech 2000;33:269–78.
[20]
Rodgers MM. Dynamic biomechanics of the normal foot and ankle during walking and running. Phys Ther 1988;68:1822–30.
[21]
An W, Rainbow MJ, Cheung RTH. Effects of Surface Inclination on the Vertical Loading Rates and Landing Pattern during the First Attempt of Barefoot Running in Habitual Shod Runners. Biomed Res Int 2015;2015:1–7. doi:10.1155/2015/240153.
[22]
Shih Y, Lin K-L, Shiang T-Y. Is the foot striking pattern more important than barefoot or shod conditions in running? Gait Posture 2013;38:490–4.
doi:10.1016/j.gaitpost.2013.01.030. [23]
Clark KP, Ryan LJ, Weyand PG. A general relationship links gait mechanics and running ground reaction forces. J Exp Biol 2016:jeb.138057. doi:10.1242/jeb.138057.
[24]
Clark KP, Ryan LJ, Weyand PG. Foot speed, foot-strike and footwear: linking gait mechanics and running ground reaction forces. J Exp Biol 2014;217:2037–40. doi:10.1242/jeb.099523.
[25]
Zadpoor AA, Nikooyan AA. The relationship between lower-extremity stress fractures and the ground reaction force: a systematic review. Clin Biomech (Bristol, Avon) 2011;26:23–8. doi:10.1016/j.clinbiomech.2010.08.005.
[26]
Novacheck T. The biomechanics of running. Gait Posture 1998;7:77–95.
[27]
Lucas-Cuevas AG, Priego Quesada JI, Giménez JV, Aparicio I, JimenezPerez I, Pérez-Soriano P. Initiating running barefoot: Effects on muscle activation and impact accelerations in habitually rearfoot shod runners. Eur J Sport Sci 2016;1391:1–8. doi:10.1080/17461391.2016.1197317.
[28]
Stief F, Böhm H, Michel K, Schwirtz A, Döderlein L. Reliability and accuracy in three-dimensional gait analysis: a comparison of two lower body protocols. J Appl Biomech 2013;29:105–11.
Figure Legends Figure 1: Kinematic data over an entire gait cycle; mean ± 1 standard deviation for runners in the barefoot (red lines) and shod conditions (black lines) (n=50) with significant differences between conditions (p<0.05, grey shaded band).
Figure 2: Kinetic data over the 50% of the gait cycle; mean ± 1 standard deviation for runners in the barefoot (red lines) and shod conditions (black lines) (n=50) with significant differences between conditions (p<0.05, grey shaded band).
Figure 3: Muscle activity data (normalized to a percentage of sprint maximum) over an entire gait cycle; mean ± 1 standard deviation for runners in the barefoot (red lines) and shod conditions (black lines) (n=50) with significant differences between conditions (p<0.05, grey shaded band)
Table 1. Discrete kinetic and spatiotemporal variables Barefoot Shod Initial rate of loading (BW/s) 92.43(97.56) 63.34(46.90)* Joint Stiffness (Nm/deg) 8.66(4.52) 13.32(6.77)** Ankle 7.27(3.91) 6.76(3.49) Knee Ground contact time (s) 0.256(0.020) 0.269(0.024)** *p<0.01; **p<0.0001 – Significant footwear condition difference.
Table 2: Summary table with respect to SPM analyses. Presented outcomes for habitual shod runners in shod and barefoot conditions Variables
Critical threshold exceeded (% of gait cycle)
Supra-threshold p-value
Critical threshold (*f)
Late Stance (18-23%)
P=0.038
F=8.581
Swing phase (74-90%)
P=0.004
Early Stance (0-6%) Mid Stance (28-38%)
P=0.032 P=0.013
Swing phase (81-100%)
P<0.001
Early Stance (2%) Early Stance (4-5%) Early Stance (5-11%)
P=0.047 P=0.035 P<0.001
F=11.737
P=0.027
F=8.820
Kinematics Knee Ankle
F=8.854
Moments Knee Ankle
F=11.661
Ground Reaction Force Vertical
Early Stance (3-10%)
S0966-6362(17)30849-4 http://dx.doi.org/10.1016/j.gaitpost.2017.08.014 GAIPOS 5762
To appear in:
Gait & Posture
Received date: Revised date: Accepted date:
17-3-2017 8-8-2017 11-8-2017
Please cite this article as: Tam Nicholas, Prins Danielle, Divekar Nikhil V, Lamberts Robert P.Biomechanical analysis of gait waveform data exploring differences between shod and barefoot running in habitually shod runners.Gait and Posture http://dx.doi.org/10.1016/j.gaitpost.2017.08.014 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Biomechanical analysis of gait waveform data exploring differences between shod and barefoot running in habitually shod runners.
Nicholas Tama, Danielle Prinsa, Nikhil V. Divekara, Robert P. Lambertsa,b.
a-
Division of Exercise Science and Sports Medicine, Department of Human Biology,
Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa. b
-Institute of Sport and Exercise Medicine, Division of Orthopaedic Surgery,
Department of Surgical Sciences, Faulty of Medicine and Health Sciences, Stellenbosch University, Tygerberg, South Africa.
Word count: 3077 Abstract count: 258 Number of Figures: 3 Number of Tables: 2
Address for correspondence: Nicholas Tam, Division of Exercise Science and Sports Medicine, Department of Human Biology Faculty of Health Sciences, University of Cape Town 3rd Floor, The Sports Science Institute of South Africa, Boundary Road, Newlands, 7725 Cape Town, South Africa [email protected]
Highlights
Differences between barefoot and shod gait also occur in swing phase. Lower limb joint co-ordination appears to differ between barefoot and shod gait Habitual shod runners exhibit earlier movement initiation when barefoot than shod Neuromuscular control in response to differing footwear requires further research. Statistical parametric mapping may provide important insight into gait where previously ignored.
Abstract: The aim of this study was to utilise one-dimensional statistical parametric mapping to compare differences between biomechanical and electromyographical waveforms in runners when running in barefoot or shod conditions. Fifty habitually shod runners were assessed during overground running at their current 10-km race running speed. Electromyography, kinematics and ground reaction forces were collected during these running trials. Joint kinetics were calculated using inverse dynamics. One-dimensional statistical parametric mapping one sample t-test was conducted to assess differences over an entire gait cycle on the variables of interest when barefoot or shod (p<0.05). Only sagittal plane differences were found between barefoot and shod conditions at the knee during late stance (18-23% of the gait cycle) and swing phase (74-90%); at the ankle early stance (0-6%), mid-stance (28-38%) and swing phase (81-100%). Differences in sagittal plane moments were also found at the ankle during early stance (2, 4-5%) and knee during early stance (5-11%). Condition differences were also found in vertical ground reaction force during early stance between (3-10%). An acute bout of barefoot running in habitual shod runners invokes temporal differences throughout the gait cycle. Specifically, a co-ordinative responses between the knee and ankle joint in the sagittal plane with a delay in the impact transient peak; onset of the knee extension and ankle plantarflexion moment in the shod compared to barefoot condition was found. This appears to affect the delay in knee extension and ankle plantarflexion during late stance. This study provides a glimpse into the co-ordination of the lower limb when running in differing footwear.
Key words: biomechanics; footwear; exercise; injury; gait.
Introduction: The statistical analysis of running gait is often restricted to the definition and extraction of discrete parameters of the gait pattern such as peaks, ranges and instances considered meaningful. Various methods have previously been employed such as principal component analysis and support vector machine, but both are pattern recognition techniques used to extract important data from large datasets [1,2]. These techniques are complex and often require large datasets to identify small changes. Recently, the use of one-dimensional statistical parametric mapping has been proposed to assess biomechanical waveform data considering the temporal nature of these data sets[3,4]. In the case of describing barefoot running gait the variables of interest largely focus on discrete loci, such as the joint kinematics at initial ground contact, which include foot strike pattern and the initial loading rate [5–7]. In addition, data often are expressed and summarized over certain periods of the running gait, for example range of motion or weight transfer during the stance phase. Based on the assumption that risk factors are most strongly correlated to the ground contact phase, most running injury research has focused on parameters measured between initial ground contact and the peak vertical ground reaction force [8].Differences in footstrike pattern at initial ground contact when running in running shoes or barefoot have been of a special interest for many researchers especially with regards to its relationship with initial loading rate [9,10]. However, recent studies have reported conflicting differences in the initial loading rate of the vertical ground reaction forced between barefoot and shod runners [5,7,11]. Although the reasoning for researching biomechanical variables between initial ground contact and peak vertical ground reaction force are meritorious, no study to date has researched if other biomechanical events outside of this period contribute to footwear condition differences. Early or late changes in joint biomechanics can potentially contribute to explaining the differences around the initial landing phase. It would be of interest to describe the periods surrounding the above-mentioned epoch and further determine the instances where differences begin and end on a temporal scale. This potentially will assist researchers in determining the period in the gait cycle that may influence these discrete events of interest. It also would be of interest to describe the differences in muscle activity that may are associated with running in the barefoot or shod condition, as peak and average amplitude differences have been found previously in habitually shod runners when running barefoot or shod [12–
14]. Where lower tibialis anterior, lateral gastrocnemius and biceps femoris activity was found when barefoot over the entire period of stance. The aim of this study is to describe the differences between barefoot and shod kinematics, kinetics and electromyography over the gait cycle in habitually shod runners using one-dimensional statistical parametric mapping (1DSPM) [3]. Specifically, we aim to investigate variables during swing that may discriminate between the two footwear conditions. It was hypothesized that kinematic differences between shod and barefoot running would be found beyond stance phase and that kinetic differences between conditions would be restricted to the this period. Methods: Participants: Fifty habitually shod (traditional cushioned shoes) runners volunteered to participate in this study. Participants were able to run 10km in <50 minutes and were injury free for six months prior to the study. Participants provided written informed consent and were fully aware of the benefits and potential risks associated with the study. The study was granted ethical approval by the Human Research Ethics Committee of the study institution and adheres the principles laid down in the Declaration of Helsinki (2013). Experimental conditions: Biomechanical testing was conducted under two different conditions, namely (1) barefoot running condition and (2) in the running shoe in which they were currently completing the most training mileage. All shod midsoles comprised of traditional ethylene vinyl acetate (EVA) cushioned shoes and were not controlled for mileage, shore count or heel-toe drop except in the case of a marketed minimalist shoe. The runners were afforded a familiarization (two lengths of the running track) in condition before performing the running trials Instrumentation: Running trials were conducted on a 40 m indoor synthetic running track. Threedimensional marker trajectories were captured using an 8-camera VICON MX motion analysis system (Oxford Metrics Ltd, Oxford, UK), sampling at 250 Hz. Ground reaction force (GRF) data were collected using a 900x600 mm force platform (AMTI, Watertown, MA, USA), sampling at 2000 Hz. Sixteen Reflective markers were attached according to a modified Helen-Hayes Marker set and the lower body PlugInGait model was applied.
Surface electromyography was measured for four lower limb muscles, namely rectus femoris (RF), biceps femoris (BF), tibialis anterior (TA) and gastrocnemius medius (MG). Prior to placement, the skin areas were prepared and two surface electrodes placed on the muscle location according to SENIAM guidelines[15] Leads and preamplifiers connected to the electrodes were secured with medical grade tape to avoid artefacts from lower limb movement during running. The transmitter unit was secured in a harness strapped to the participant’s back and data sampled at 2000 Hz (Noraxon 2400T G2, Noraxon, USA). Procedures: Participants completed 6 clean overground running trials in each footwear condition in a randomized order, with no instruction to running style. The speed of overground running trials was set based on the participant’s current 10km performance pace (within a month). Trials were accepted if the velocity was within ± 5% of the target speed. During these runs, synchronised collection of marker motion, force platform and electromyography measurements were obtained, a successful trial was defined as one within the specified velocity range, where all markers were in view of the cameras and there was no visual evidence of force platform targeting. After completing running trials at the designated speeds, the runners then completed three maximal sprints down the 60 m runway in the shod condition. Data analysis: Marker trajectory and force platform data were filtered using a low-pass fourth-order Butterworth filter with a cut-off frequency at 20 and 100 Hz respectively. For each trial, one complete stance phase of the gait cycle was analysed. Three-dimensional lower extremity joint angles and net resultant joint moments using a Newton-Euler inverse dynamics approach were calculated [16,17]. Three-dimensional joint moments were expressed as external moments normalized to body mass (Nmkg-1). Raw digital EMG signal of both sub-maximal and sprint trials were processed as previously described [18], through rectification and root mean squared smoothing window at 50 ms. Subsequently, sub-maximal trials were normalized to peak activity from the appropriate trial. The data for each participant’s right limb were averaged over 6 trials for each condition. Sagittal, and frontal plane ankle and knee angles (degrees) and moments (Nmkg-1) are reported. Specifically, all muscle activity, kinematic and kinetic data are represented as waveforms that changed continuously throughout the gait cycle and were defined with 101 data points, one for each percentage of the cycle.
Statistical analysis Data were screened for normality of distribution using a Shapiro-Wilk’s Normality test. Differences in initial loading rate; ground contact time; knee and ankle joint stiffness were assessed using t-tests or non-parametric Wilcoxon signed rank test. To detect significant differences between the kinematic and kinetic waveforms in an objective way, 1DSPM was employed [3]. The barefoot and shod conditions were compared to each other using an one-way ANOVA (SPM{f}) All 1DSPM analyses were implemented using the open-source 1DSPM code (v.M0.1, www.spm1d.org) in Matlab (R2014a, 8.3.0.532, Mathworks Inc., Natick, MA, USA). Significance was set at p<0.05. Results: The characteristics of the group of runners were age: 30.3 ± 7.4 years, mass: 71.5 ± 11.1 kg, height: 1.8 ± 1.0 m and recent 10-km time: 48.3 ± 7.8 min. The mean running speed during over-ground running trials was 3.5 ± 0.5 ms-1 and no differences in running speed between the footwear conditions were found (3.5 ± 0.4 ms-1 vs. 3.5 ± 0.6 ms-1 for shod and barefoot, respectively). Both vertical initial loading rate and ankle stiffness were greater in the barefoot running condition than in the shod running conditions, while ground contact time was shorter when running barefoot (Table 1). Although no waveform difference were found in the frontal and transfers planes between the barefoot or shod condition, waveform differences were found in the sagittal plane (see Table 2 and Figure 1). Kinematics A notably reduced knee flexion was found between 18-23% of the gait cycle (F=8.581; P=0.038) in the barefoot condition than when running shod. In line with this an earlier knee extension response during late swing (74-90% of the gait cycle) was seen in the barefoot condition (F=8.581; P=0.004). In addition, reduced ankle dorsiflexion was observed in the barefoot running condition during the 0-6% of the gait cycle (F=8.854; P=0.032) (Table 2 and Figure 1). During weight transfer ankle plantarflexion between 28-38% of the gait cycle (F=8.854; P=0.013) and during late swing at 81-100% of the gait cycle (F=8.854; P<0.001) was greater in the barefoot when compared to shod condition. Forces and moments
Differences in the vertical ground reaction force were found at early onset of stance where the rise in force when running barefoot was earlier than shod at 3-11% of the gait cycle (F=8.820; P=0.027) (Table 2 and Figure 2). Greater barefoot knee extension moments were found at 2 and 4-5% of the gait cycle (F=11.737; P=0.47 and P=0.035 respectively). Alongside this, greater barefoot ankle plantarflexion moments were found between 5-11% of the gait cycle (F=11.661; P<0.001). Muscle activity No differences in muscle activity waveforms were found between the two conditions (Figure 3). Discussion: The novelty of this study is that differences between barefoot and shod gait have been found throughout the gait cycle including swing phase using statistical parametric mapping. In addition, the analysis technique was able to identify the temporal periods of the gait cycle. This expands our knowledge of changes in gait associated with different footwear that may potentially impact on the understanding of function, injury and performance, as previous footwear research has largely focused on ground contact phases. Swing Phase Swing phase of gait is often overlooked, as it does not directly assess the interaction between the moving body and it’s surrounding environment [19]. Hence, the ground contact phase of gait has mainly been studied as a consequence of the main epoch where large forces are experienced during gait. However, during terminal swing of the gait cycle (74-90%), barefoot knee extension started earlier and was greater than when running in the shod condition. In addition, it appears that peak knee flexion during swing may have been slightly reduced when running barefoot, however this difference was non-significant. Coinciding with the earlier knee extension response, we found that the ankle during barefoot running remains in a neutral position during terminal swing phase of the gait cycle (81-100%), while the ankle is dorsiflexed during this phase when running in a shod condition. These findings reflect an adaptive barefoot running style, which allows the barefoot runner to have a more mid-foot landing pattern. In the shod condition, more knee flexion and ankle dorsiflexion is needed to allow adequate foot ground clearance and a heel strike landing pattern [20]. The more mid-foot landing pattern in the barefoot running condition is also confirmed by the early phase of the stance phase (0-6% of the gait cycle|) when the ankle remains neutral in the barefoot condition and is still in maximal
dorsiflexion in the shod running condition. These findings support a mid/forefoot landing in the barefoot running condition and a heelstrike footstrike pattern in the shod runnning condition [21,22]. In addition, this footwear difference also illustrates the influence of shoe cushioning/construction on gait as the ankle is required to preemptively dorsiflex prior ground contact in order to clear the ground during swing as a result of the medium (footwear cushioning) between the foot and the ground. Ground Contact Phase The first difference found at ground contact was ankle plantar-dorsiflexion from 0-6% of the gait cycle. The barefoot condition remains in neutral from swing phase and the shod condition in dorsiflexion throughout initial ground contact where both conditions merge at foot flat phase and initial impact peak. Prior to ground contact the barefoot condition remains in a neutral angle during terminal swing and this persists at initial ground contact. This observation at initial stance phase has previously been documented in previous research assessing discrete variables [7,19]. However, contrary to previous research, this study found that temporal differences continue over a quarter of the gait cycle. This difference is a result of runners striving for a mid/forefoot landing when barefoot as this would reduce the discomfort of a rear foot landing [19]. We can postulate that runners alter their foot strike pattern as a result of the cushioning of the shoe giving them the ease to land more rearfoot [19,22]. Concurrently, the barefoot vertical ground reaction force produced a earlier but lower force at 3-10% of the gait cycle, while a later but higher ground reaction force was seen in the shod running condition. Ankle plantarflexion influences the vertical ground reaction force since a mid/forefoot landing creates a greater landing surface [6,23] and consequently the impact is distributed across a larger surface area resulting in a rise in peak. Further, this also informs us of the neuromuscular anticipation and preparation required for landing in the absence of footwear cushioning, despite not finding any muscle activity differences between the two conditions. As previously stated an earlier rise in the vertical ground reaction force was found in the barefoot condition, this was associated with a greater initial loading rate too. A greater initial loading rate has previously been linked to risk factors for tibial stress factors because recent evidence has suggest that the shank and foot mass interaction largely influence the initial ground contact kinetics and up to impact transient peak [23–25]. The impact transient peaks between the two conditions are occur at differing times of the gait cycle and maybe different in magnitude. Thus, caution may still be advised on acute transition to barefoot running as this rate of
force development during early stance may increase injury risk in the shank-foot complex. Differences in ankle sagittal kinematics and vertical ground reaction force intersect and proceed with a greater barefoot knee extension moment and ankle plantarflexion moment between 2; 3-5% and 5-11% of the gait cycle, respectively. A larger barefoot knee extension moment was observed prior to the difference in ankle plantarflexion moment. These findings inform us of the kinematic response to the lower limb landing mechanism and suggest the influence of neuromuscular control in the absence of footwear. This earlier and greater barefoot knee extension moment is an indicator of the eccentric action of the quadriceps in response to absent footwear cushioning and neutral ankle flexion angle at ground contact. Subsequently, the onset of shod ankle plantarflexion moment was delayed as a consequence of differing kinematic landing strategies because the shod ankle enters stance already in dorsiflexion. Therefore, the need to increase the plantarflexion moment is reduced or delayed when shod compared to the barefoot condition. The onset of the barefoot plantarflexion moment typically occurs between 5-10% of the gait cycle [26]. Following this difference in ankle flexion during early stance, greater shod knee flexion between 18-23% of the gait cycle when compared to the barefoot condition. This finding suggests that in the shod condition knee flexion is both greater and maintained through weight transfer of stance. This finding suggests a joint kinematic response to weight acceptance and transfer in the presence or lack of footwear cushioning on the joint mechanics [7]. Interestingly, it appears that this increased knee flexion in the shod condition maybe a response to the greater ankle stiffness as result of the greater dorsiflexion of the ankle at the beginning of stance that reduces ankle range of motion. Subsequently, this may also impact ankle plantarflexion during late stance and towards toe-off as a difference was found between 28-38% of the gait cycle. This difference appears to be a result of delayed plantarflexion in the shod condition, as a consequence of delayed knee extension from peak flexion during weight transfer, as peak knee flexion during stance appears to be sustained and is greater (difference between 18-23% of the gait cycle) than barefoot in the shod condition. This difference also reflects a delayed response in ankle plantarflexion in the shod condition, with the barefoot condition obtaining an earlier peak plantarflexion angle at toe-off prior to toe-off.
Differences in ground contact time were found although all trials were performed at the same speed. Whether or not this may impact on the differences found in this study are unknown, however, the entire gait cycle was normalised to 101 points for ease of interpretation and reference. Further, no differences were found between conditions in muscle activity. Previous researchers have found amplitude differences over the differing phases of gait in barefoot and shod conditions [12,13,22,27]. However, many of these studies focus on differing footfall patterns and imposing differing gait patterns. Thus, the instinctive adaptation is often ignored or not of interest, whereas this study provided no such feedback and segregation of footfall patterns. Ultimately, the high variability of the muscle activity data did not allow this technique to infer whether the data sets were indeed different from random. Further, an important consideration regarding this study is that the commonly used PlugInGait model has been found to reproduce highly variable transverse plane angles. This limits determining clinically relevant changes/differences between footwear conditions [28], thus the use of a different model may reveal differences not found in this study. This study remains relevant as it suggests to researchers to refrain from over simplifying gait data and to appreciate the entire waveform and further supports analyses like 1DSPM to better interpret and understand changes in gait. Conclusion: This study provides insight into terminal swing, ground contact preparation phase and loading rate dynamics when running in a shod or barefoot condition. Differences between barefoot and shod were mainly seen in the knee and ankle joint in the sagittal plane and landing forces. Notable footwear differences were seen during late swing and early landing phase. Barefoot running was characterised by a mid-/forefootstrike pattern, whereby preparation starts during terminal swing phase exhibited by an earlier knee extension response and a more neutral ankle flexion position during the final part of the swing and early stance phase. These subtle differences seem to reflect complex motor control that adapts gait in different footwear conditions. However, as running in different conditions alters gait, runners should take slowly introduce new running styles rather than drastically swop over. Conflicts of interest: No conflict of interest
References: [1]
Trudeau MB, von Tscharner V, Vienneau J, Hoerzer S, Nigg BM. Assessing Footwear Effects from Principal Features of Plantar Loading during Running. Med Sci Sports Exerc 2015;47:1988–96. doi:10.1249/MSS.0000000000000615.
[2]
Astephen JL, Deluzio KJ. A multivariate gait data analysis technique: application to knee osteoarthritis. Proc Inst Mech Eng H 2004;218:271–9.
[3]
Pataky TC, Vanrenterghem J, Robinson MA. Zero- vs. one-dimensional, parametric vs. non-parametric, and confidence interval vs. hypothesis testing procedures in one-dimensional biomechanical trajectory analysis. J Biomech 2015;48:1277–85. doi:10.1016/j.jbiomech.2015.02.051.
[4]
Nedergaard NJ, Robinson MA, Eusterwiemann E, Drust B, Lisboa PJ, Vanrenterghem J. The Relationship Between Whole-Body External Loading and Body-Worn Accelerometry During Team Sports Movements. Int J Sports Physiol Perform 2016:1–23. doi:10.1123/ijspp.2015-0712.
[5]
Willy RW, Davis IS. Kinematic and kinetic comparison of running in standard and minimalist shoes. Med Sci Sports Exerc 2014;46:318–23. doi:10.1249/MSS.0b013e3182a595d2.
[6]
Lieberman DE, Venkadesan M, Werbel WA, Daoud AI, D’Andrea S, Davis IS, et al. Foot strike patterns and collision forces in habitually barefoot versus shod runners. Nature 2010;463:531–5. doi:10.1038/nature08723.
[7]
Tam N, Astephen Wilson JL, Coetzee DR, van Pletsen L, Tucker R. Loading rate increases during barefoot running in habitually shod runners: Individual responses to an unfamiliar condition. Gait Posture 2016;46:47–52. doi:10.1016/j.gaitpost.2016.02.013.
[8]
Tam N, Astephen Wilson JL, Noakes TD, Tucker R. Barefoot running: an evaluation of current hypothesis, future research and clinical applications. Br J Sports Med 2014;48:349–55. doi:10.1136/bjsports-2013-092404.
[9]
Nigg BM. The role of impact forces and foot pronation: a new paradigm. Clin J Sport Med 2001;11:2–9.
[10]
Davis IS, Bowser BJ, Mullineaux DR. Greater vertical impact loading in female runners with medically diagnosed injuries: a prospective investigation. Br J Sports Med 2016;50:887–92. doi:10.1136/bjsports-2015-094579.
[11]
Rice HM, Jamison ST, Davis IS. Footwear Matters: Influence of Footwear and Foot Strike on Loadrates During Running. Med Sci Sports Exerc 2016:2462–8. doi:10.1249/MSS.0000000000001030.
[12]
Ervilha UF, Mochizuki L, Jr AF. Are muscle activation patterns altered during shod and barefoot running with a forefoot footfall pattern ? J Sports Sci 2016;0:1–7. doi:10.1080/02640414.2016.1231410.
[13]
Tam N, Tucker R, Astephen Wilson JL. Individual Responses to a Barefoot Running Program: Insight Into Risk of Injury. Am J Sports Med 2016;44:777– 84. doi:10.1177/0363546515620584.
[14]
Snow NJ, Basset FA, Byrne J. An Acute Bout of Barefoot Running Alters Lower-limb Muscle Activation for Minimalist Shoe Users. Int J Sports Med 2016;37:382–7. doi:10.1055/s-0035-1565140.
[15]
Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 2000;10:361–74.
[16]
Winter DA. Overall principle of lower limb support during stance phase of gait. J Biomech 1980;13:923–7. doi:10.1016/0021-9290(80)90162-1.
[17]
Grood ES, Suntay WJ. A joint coordinate system for the clinical description of three-dimensional motions: application to the knee. J Biomech Eng 1983;105:136–44. doi:6865355.
[18]
Albertus-Kajee Y, Tucker R, Derman W, Lamberts RP, Lambert MI. Alternative methods of normalising EMG during running. J Electromyogr Kinesiol 2011;21:579–86. doi:10.1016/j.jelekin.2011.03.009.
[19]
DeWit B, De Clercq D, Aerts P. Biomechanical analysis of the stance phase during barefoot and shod running. J Biomech 2000;33:269–78.
[20]
Rodgers MM. Dynamic biomechanics of the normal foot and ankle during walking and running. Phys Ther 1988;68:1822–30.
[21]
An W, Rainbow MJ, Cheung RTH. Effects of Surface Inclination on the Vertical Loading Rates and Landing Pattern during the First Attempt of Barefoot Running in Habitual Shod Runners. Biomed Res Int 2015;2015:1–7. doi:10.1155/2015/240153.
[22]
Shih Y, Lin K-L, Shiang T-Y. Is the foot striking pattern more important than barefoot or shod conditions in running? Gait Posture 2013;38:490–4.
doi:10.1016/j.gaitpost.2013.01.030. [23]
Clark KP, Ryan LJ, Weyand PG. A general relationship links gait mechanics and running ground reaction forces. J Exp Biol 2016:jeb.138057. doi:10.1242/jeb.138057.
[24]
Clark KP, Ryan LJ, Weyand PG. Foot speed, foot-strike and footwear: linking gait mechanics and running ground reaction forces. J Exp Biol 2014;217:2037–40. doi:10.1242/jeb.099523.
[25]
Zadpoor AA, Nikooyan AA. The relationship between lower-extremity stress fractures and the ground reaction force: a systematic review. Clin Biomech (Bristol, Avon) 2011;26:23–8. doi:10.1016/j.clinbiomech.2010.08.005.
[26]
Novacheck T. The biomechanics of running. Gait Posture 1998;7:77–95.
[27]
Lucas-Cuevas AG, Priego Quesada JI, Giménez JV, Aparicio I, JimenezPerez I, Pérez-Soriano P. Initiating running barefoot: Effects on muscle activation and impact accelerations in habitually rearfoot shod runners. Eur J Sport Sci 2016;1391:1–8. doi:10.1080/17461391.2016.1197317.
[28]
Stief F, Böhm H, Michel K, Schwirtz A, Döderlein L. Reliability and accuracy in three-dimensional gait analysis: a comparison of two lower body protocols. J Appl Biomech 2013;29:105–11.
Figure Legends Figure 1: Kinematic data over an entire gait cycle; mean ± 1 standard deviation for runners in the barefoot (red lines) and shod conditions (black lines) (n=50) with significant differences between conditions (p<0.05, grey shaded band).
Figure 2: Kinetic data over the 50% of the gait cycle; mean ± 1 standard deviation for runners in the barefoot (red lines) and shod conditions (black lines) (n=50) with significant differences between conditions (p<0.05, grey shaded band).
Figure 3: Muscle activity data (normalized to a percentage of sprint maximum) over an entire gait cycle; mean ± 1 standard deviation for runners in the barefoot (red lines) and shod conditions (black lines) (n=50) with significant differences between conditions (p<0.05, grey shaded band)
Table 1. Discrete kinetic and spatiotemporal variables Barefoot Shod Initial rate of loading (BW/s) 92.43(97.56) 63.34(46.90)* Joint Stiffness (Nm/deg) 8.66(4.52) 13.32(6.77)** Ankle 7.27(3.91) 6.76(3.49) Knee Ground contact time (s) 0.256(0.020) 0.269(0.024)** *p<0.01; **p<0.0001 – Significant footwear condition difference.
Table 2: Summary table with respect to SPM analyses. Presented outcomes for habitual shod runners in shod and barefoot conditions Variables
Critical threshold exceeded (% of gait cycle)
Supra-threshold p-value
Critical threshold (*f)
Late Stance (18-23%)
P=0.038
F=8.581
Swing phase (74-90%)
P=0.004
Early Stance (0-6%) Mid Stance (28-38%)
P=0.032 P=0.013
Swing phase (81-100%)
P<0.001
Early Stance (2%) Early Stance (4-5%) Early Stance (5-11%)
P=0.047 P=0.035 P<0.001
F=11.737
P=0.027
F=8.820
Kinematics Knee Ankle
F=8.854
Moments Knee Ankle
F=11.661
Ground Reaction Force Vertical
Early Stance (3-10%)