Can m inimal running shoes im itate barefoot heel-toe running patterns? A comparison of lower leg kinematics

Tobias Hein,Stefan Grau,*

aDepartment of Food and Nutrition,and Sport Science,University of Gothenburg,Go¨teborg 40530,Sweden

bMedical Clinic,Department of Sports Medicine,University of Tuebingen,Tuebingen 72076,Germany

Original article

Can m inimal running shoes im itate barefoot heel-toe running patterns? A comparison of lower leg kinematics

Tobias Heina,b,Stefan Graua,b,*

aDepartment of Food and Nutrition,and Sport Science,University of Gothenburg,Go¨teborg 40530,Sweden

bMedical Clinic,Department of Sports Medicine,University of Tuebingen,Tuebingen 72076,Germany

Background:Numerous studies about the interaction between footwear(and barefoot)and kinematic and kinetic outcomes have been published over the last few years.Recentstudies however lead to the conclusion that the assumed interactions depend mainly on the subjects’experience of barefoot(BF)walking/running,the preferred running strike pattern,the speed,the hardness of the surface,the thickness of the midsole material, and the runners’level of ability.The aim of the present study was to investigate lower leg kinematics of BF running and running in m inimal running shoes(MRS)to assess comparability of BF kinematics in both conditions.To systematically compare both conditions we monitored the influencing variables described above in our measurement setup.We hypothesized that running in MRS does not alter lower leg kinematics compared to BF running.

Methods:Thirty-seven subjects,injury-free and active in sports,ran BF on an EVA foam runway,and also ran shod wearing Nike Free 3.0 on a tartan indoor track.Lower-leg 3D kinematics was measured to quantify rearfootand ankle movements.Skin markers were used in both shod and BF running.

Results:All runners revealed rearfoot strike pattern when running barefoot.Differences between BF and MRS running occurred particularly during the initialstance phase of running,both in the sagittaland the frontalplanes.BF running revealed a flatter footplacement,a more plantar flexed ankle joint and less inverted rearfoot at touchdown compared to MRS running.

Conclusion:BF running does not change the landing automatically to forefoot running,especially after a systematic exclusion of surface and other influencing factors.The Nike Free 3.0 m im ics some BF features.Nevertheless,changes in design of the Nike Free should be considered in order to m im ic BF movement even more closely.

CopyrightⒸ2014,Shanghai University of Sport.Production and hosting by Elsevier B.V.All rights reserved.

3D-kinematics;Ankle;Barefoot;M inimal running shoes;Rearfoot;Running

1.Introduction

Advantages and disadvantages of barefoot(BF)running have been of major interest for numerous years,scientifically as wellas in the running population.As a consequence of this, there have been numerous concepts and products on the market thatm im ic specific aspects of BF movement,shape,or feeling,“suggesting that some of the perceived advantages of barefoot running are transferred into a shod condition”.1Scientifically,publications and discussions about advantages and disadvantages of BF running increased tremendously after a publication by Lieberman et al.2inNature.

Numerous studies about the interaction between shod and BF kinematic and kinetic outcomes have been published over the last few years and described by Nigg1and Nigg and Enders.3Mostof these studies were based on the comparison of running in traditional running shoes(TRS)and BF running. Recent studies however lead to the conclusion that the assumed interactions depend mainly on the subjects’experience w ith BF walking/running,2,4—8the preferred running strike pattern,9,10the speed,11the hardness of the surface,12the thickness of the midsole material,13and the runners’level.10A few studies4—7have already included m inimal running shoes (MRS)into their setup.

To systematically analyze suggested“barefoot features”in given MRS and compare w ith the BF situation,it is necessary to take the above-mentioned criteria into account.Therefore, studies should monitor the subjects’experience in BF walking/ running(unexperienced or experienced),the preferred running strike pattern(rearfoot,m idfoot,forefoot),the running speed (typical running speed,depending on runners’level and gender),the hardness of surface(hardness of BF running surface comparable to m idsole hardness of MRS),the thickness of midsole material(one thickness)and the subject’s athletic level(recreational,elite).Further,skin mounted markers should be used14as shoe-mounted markers are not adequate to assess the in-shoe foot motion,and consequently overestimate its realmotion.Although these results have been shown for TRS w ith stiff heel counters,the flexible heel counter of MRS m ight have an even greater influence on the resultant rearfoot and ankle kinematics.

The aim of the present study was to investigate lower leg kinematics of BF running and running in MRS(Nike Free 3.0; Nike Inc.,Beaverton,OR,USA)to assess comparability of BF kinematics in both conditions.Furthermore,we aimed to find out if foot strike characteristics remained the same after monitoring the influencing variables described above in our measurement setup.We hypothesized that running in MRS does not alter lower leg kinematics compared to BF running and that foot strike pattern remained the same in both conditions.

2.M aterials and m ethods

2.1.Subjects

Overall,37 subjects were included in the study(Table 1). A ll subjects were rearfoot strikers(visually inspected beforehand while running in theirown TRS),free ofany injury forat least 6 months prior to the study,recreational athletes (different sports)and aged between 18 and 55 years.None of the subjects had a history of or experience w ith BF running. The study complies w ith the Declaration of Helsinki,and all subjects signed a w ritten consent form prior to the testing procedures.

2.2.Experimental setup

Three-dimensional kinematics was recorded w ith a sixcamera infrared system(ViconPeak,MCam,M 1;Oxford, UK)ata sampling frequency of250 Hz.All runners ran BF on a 20-m EVA foam runway(shore hardness approx.40),and shod wearing Nike Free 3.0(shore hardness approx.40)on a 20-m tartan indoor track.The height of the foam runway was 10 mm,comparable to the m idsole/outsole heel height of MRS.The order of running conditions was randomized.Prior to the recorded measurements,sufficient time was allowed for the subjects to familiarize themselves w ith the laboratory setup and to get used to the running speed and surface to enable an individual running style.A ll subjects ran w ith a controlled running speed of 11 km/h monitored using a photoelectric barrier,and a running speed between 10.5 km/h and 11.5 km/h was accepted.The test speed of 11 km/h was chosen as this is an average running speed in recreational athletes,both for men and women.Touch-down was visually inspected to find out if subjects landed on the rearfoot or on the m id/forefoot.

Eighteen markers were placed on each subject according to the recommendations of the International Society of Biomechanics,15marking both shanks(medial and lateral tibia plateau,tibial tuberosity,medial tibial crest,lateral and medial malleoli),the foot(lateral,medial,and posterior calcaneus),and the hallux.Rearfootmarkers were screwed to a short thread(～1 cm)and screw sockets were attached to custom ized flexible plastic disks placed on the calcaneus to ensure their visibility and identical placement for both BF and shod conditions(Fig.1)and to ensure a good fi t of the markers w ith respect to the foot.Joint excursions were quantified by calculating Cardan angles according to So¨derkvist and Wedin16w ith the foot segment rotating w ith respect to the shank segment(ankle dorsiflexion/plantarflexion,rearfoot inversion/eversion),or w ith respect to the global coordinate system(tibial rotation,sagittal ankle,and frontal rearfoot motion).Further,the fi rst rotation was computed around the sagittal axis(dorsiflexion/plantarflexion),the second rotation around the frontal axis(inversion/eversion)and finally,the third rotation was computed around the transversal axis(external/internal rotation).Forthe subsequentanalysis,stance phase was detected according to Maiwald et al.17and subsequently normalized to 100 data points which equal 100%of stance phase(%SP);sw ing phase was neglected.Mean joint excursion curves and discrete variables were based on 10 valid trials for each subject.

Table 1 Overview of all subjects included in the study(mean±SD).

Fig.1.Placement of skin-mounted markers on the rearfoot and of external shoe markers(A).Skin-mounted markers for the rearfoot segment(B).

The follow ing joint motions were calculated:tibial external/internal rotation(TER,TIR),ankle dorsiflexion/ plantarflexion(ADF,APF),rearfoot inversion/eversion(RFIN, RFEV)as well as frontal rearfoot(RFGIN,RFGEV)and sagittal ankle motion(ASAG)w ith respect to the global coordinate system,respectively.

Discrete variables were:

·Initial joint excursion at touchdown(°)for tibial external rotation(TERinit),ankle dorsiflexion(ADFinit),rearfoot inversion(RFINinit),frontal rearfoot motion(RFGINinit), and sagittal ankle motion(ASAGinit);

·Maximal joint excursions(°)and tim ing(%SP)for tibial internal rotation(TIRmax,t TIRmax),ankle dorsiflexion (ADFmax,t ADFmax),rearfoot eversion (RFEVmax, t RFEVmax),and frontal rearfoot motion(RFGEVmax, t RFGEVmax);

·Maximal joint excursions(°)for tibial external rotation (TERmax),ankle plantarflexion(APFmax),rearfoot inversion(RFINmax),and frontal rearfoot motion(RFGINmax);

·Ranges of motion(°)for internal and external tibial rotation(TIRROM,TERROM),ankle dorsi/plantarflexion (ADFROM,APFROM),rearfoot eversion/inversion(RFEVROM,RFINROM),and frontal rearfoot motion(RFGEVROM, RFGINROM).

2.3.Shoes

Nike Free 3.0 served as test shoe in sizes US6(24 cm)—US12(30 cm).To record the markers placed directly on the rearfoot,three small w indows(lateral,medial,and posterior) were cut into the heel counter(Fig.1).Three markers were placed on the sole around the heel(lateral,medial,and posterior)to determ ine touchdown,and one marker was located on the tip of the shoe to calculate toe-off.Nike Free 3.0 was chosen as MRS due to the follow ing proposed“barefoot features”:extreme flexibility of complete m idsole and outsole in running and medio-lateral direction,alignment of m idsole/ outsole squares adjusted to center of pressure path in BF running,low midsole heightand flexible textile uppermaterial (whole upper material,no stiff heel counter).

2.4.Statistical analysis

The analysis of kinematic data was based on a randomized side for each subject to avoid bias,such as dom inantvs.nondom inant leg if only one side were consistently evaluated.A comparison of dom inant or non-dom inant leg only was not applicable as not all subjects were able to define which side was dom inant.Thus,18 right and 19 left sides were included in the subsequentanalysis.Continuous means of 10 trials and 95%confidence intervals(CI)were computed for both conditions and for each joint motion to present kinematic data. Non-overlapping CIindicates a significantdifference between BF and MRS conditions.

Group means,standard deviations(SD),medians and the upper and lower lim its of the 95%CI were calculated, dependentttests were conducted to analyze differences between BF and MRS conditions.The level of significance was set atα=0.05.Since 25 variables were analyzed,25ttests were performed and the test level was adjusted according to the Bonferroni procedure18top＜0.002(0.05/25=0.002). Statistical analyses were carried out using JMP(version 10, SAS Institute Inc.,Cary,NC,USA).

3.Results

A ll participants revealed the same foot strike pattern in BF and MRS.The results of lower leg kinematics of BF running and running in MRS(Nike Free 3.0)to assess comparability of BF kinematics in both conditions are shown in Table 2 and Fig.2.For transversal tibia motion,running in MRS showed a later t TIRmaxand a decreased TERmaxcompared to BF.W ith regard to sagittalankle movement,MRS was accompanied by greater ADFinitand ADFmaxas well as a later t ADFmax.The greater APFROMfor MRS was derived subsequently from the differences in ADFmaxand equal amounts of APFmax.Frontalrearfoot kinematics revealed a more inverted rearfoot at touchdown(RFINinit),a later t RFEVmaxand an increased RFEVROMin the MRS condition.The increased RFEVROMwas a consequence of increased RFINinitand equalamounts of RFEVmax.

Table 2 Three dimensional lower leg kinematics for running barefoot(BF)and shod,wearing m inimal running shoes(MRS).Displayed are group means±SD,group medians,upper and lower lim its of the 95%confidence intervals(CI).

With respect to the global system, the rearfoot segment reached maximal eversion at a later point in time (t RFGEVmax)and revealed a greater RFGINROMfor MRS compared to BF.Finally,ASAGinitdecreased by 9°for BF compared to MRS.

4.Discussion

One aim of the present study was to investigate lower leg kinematics in BF running and running in an MRS(Nike Free 3.0)to assess comparability of BF kinematics in both conditions.To systematically compare both conditions,we monitored influencing variables such as BF experience,preferred running strike pattern,speed,hardness and height of surface, and the athletes’level.Finally,we applied skin-mounted markers in both conditions to avoid bias due to marker placement.We hypothesized that running in an MRS does not alter lower leg kinematics compared to BF running.In summary,many differences in lower leg kinematics were found between the two conditions concerning transversal tibia, sagittal ankle and frontal rearfoot kinematics.Especially initial touch-down in frontal rearfoot and sagittal ankle kinematics were different between the two different conditions. Frontal rearfootkinematics showed a more inverted rearfootat touchdown and throughout the initial contact phase for MRS, whereas sagittal ankle kinematics showed a more dorsi flexed ankle joint and a higher sagittal touchdown angle in MRS. Additionally,tim ing of several maximal joint excursions for tibia and ankle in allplanes were delayed in MRS compared to BF.Thus our hypothesis had to be rejected.

The results of our study have to be discussed mainly in two directions:fi rst,a comparison w ith the existing literature and second,rating of findings w ith regard to proposed“barefoot features”.

Fig.2.Continuous jointexcursion curves for barefoot running(BF,gray)and running w ith m inimal running shoes(MRS,white):tibial external/internal rotation (A),ankle dorsi/plantarflexion(B),rearfoot inversion/eversion(C),and frontal rearfoot inversion/eversion with respect to the global system(D).Displayed are means and 95%confidence bands over the entire stance phase.

Differences in frontal rearfoot and sagittal ankle joint excursion at touchdown,or slightly prior to touchdown,have been reported in different studies.4—7,13,19,20Additionally Bonacci et al.,4Sinclair et al.,6and TenBroek et al.13found a less dorsi flexed ankle at touchdown for BF compared to the MRS condition and an even more increased dorsiflexion in TRS.Our data(absolute values)correspond very wellw ith the data of Bonacci et al.,4and the absolute values of Sinclair etal.6and TenBroek etal.13differ slightly whereas the relative differences align well w ith our data.The studies by Bonacci etal.4and Sinclair etal.6used the same MRS thatwas used in our study.Different results were found in Squadrone and Gallozzi’s article,7where no differences of sagittal ankle kinematics at touchdown were reported between BF and MRS, butbetween TRS and both other conditions.A reason for the contradictory result could be the slightly different method of data evaluation at touchdown phase.Thus,Squadrone and Gallozzi7analyzed sagittal ankle kinematics 15 ms prior to touchdown whereas the other studies including the current study analyzed jointexcursion at touchdown.Further,the use of different MRS conditions(Nike Free 3.0 and Vibram FiveFingers™(Vibram,A lbizzate,Italy))m ight also influence the results,since the Vibram FiveFingers™is basically a sock thatcovers the foot.Lastly,although Paquette etal.5stated that there was no difference in strike patterns at touchdown for kinematics between BF and MRS w ith regard to TRS,the study results were not comparable w ith ours since Paquette et al.5evaluated pressure data instead of kinematic data.On the other hand,Bishop et al.19also reported a decrease in ankle joint dorsiflexion at touchdown for BF compared w ith TRS,whereas the absolute values corresponded well with the Sinclair etal.data.6Sim ilar results were found in the study by De W it et al.20in 2000.In summary,although influencing factors were not or hardly verified in many of the described studies and although the applied methods(2Dvs.3D)and calculation routines differ between the different approaches, we assume that our data support the current research even under a strictly monitored measurement setup.As we did not include a TRS condition in our study design,based on the findings from Sinclair etal.6and Squadrone and Gallozzi,7we would suppose sagittal ankle joint excursions wearing the Nike Free 3.0 to be between BF and TRS,whereas we presume the Vibram FiveFingers™MRS to be closer to BF.

Our results present a more inverted rearfoot at touchdown and throughout initial contact phase for MRS,which seems to concur w ith the studies by Bonacci et al.4and TenBroek et al.13We assume an increased inversion of the rearfoot at touchdown to be a consequence of a more dorsiflexed ankle joint,since the tibialis anterior muscle,as main dorsiflexor muscle,also leads to increased inversion due to its insertion at the medialside of the foot.The study by Sinclair etal.6did not report any differences between BF and MRS,which is in contrast to our findings.We would presume that these contradicting results are due to the differentmarkersets at the foot and consequently due to the differentcalculation methods usedto quantify frontal rearfoot motion.No differences between BF and MRS towards TRS were reported in either paper.No comparing data could be derived from Paquette et al.5or Squadrone and Gallozzi.7Further,no information about the inversion of the rearfootat touchdown was found in the papers by Bishop etal.19and De Witetal.20As we did not include a TRS condition in our study design,based on the above described findings we would assume that frontal ankle kinematics in wearing the Nike Free 3.0 m ight be closer to TRS than to BF(although the absolute differences between BF and MRS might be smaller).

No other studies were found that reported a delay in the tim ing of several maximal joint excursions in MRS compared to BF or compared w ith TRS for tibia,ankle,and rearfoot kinematics.From our point of view,these results m ight be explainable by the extreme flexibility of the m idsole squares due to the numerous flex grooves in the Nike Free 3.0 in medio-lateral and anterior-posterior directions,whereas the foam runway,though comparable in hardness and height, showed consistent material properties w ith no flex grooves. Thus,this resultm ightbe setup-specific.

The differencesin TERROMand RFGINROMbetween BF and MRS during the second halfof the stance phase also seem to be setup-specific.We would speculate that the EVA foam cannot provide sufficient friction for a straight push-off phase.Thus, slighttorsion under the forefootmay lead to an outward rotation of the rearfoot and finally to an increased inversion of the rearfootduring BF running.In contrast,the flexible rubbersole of the Nike Free 3.0 and the tartan surface produce enough friction to enable a straightpush-off phase for MRS running.

Foot strike pattern at touchdown did not change in our study.This would indirectly support the findings of Gruber et al.12who reported different foot strike patterns when the same subjects were running on different hard surfaces.Thus, barefoot running does not change the landing automatically to forefoot running.Besides the hardness of the surface,other factors like speed and subjects’experience w ith BF running m ight attribute more to a change the landing pattern from rearfoot to forefoot orvice versa.

We are aware of several lim itations to this study which must be considered.One major lim itation is the m issing TRS condition,meaning that kinematic data from BF,MRS,and TRS conditions could not be obtained and compared w ith recent literature.Since ground reaction forces were notmeasured and inverse dynam ics not calculated,the authors cannot comment upon the occurrence of an impact peak at touchdown(to quantitatively determ ine strike pattern at touchdown),the differences in impact peaks,loading rates,or resulting jointmoments between the two conditions.Quantifying and evaluating hip and knee kinematics would have been beneficial for the currentstudy since the demonstrated differences in lower limb mechanics m ightalter hip and knee kinematics.1

5.Conclusion

The current study revealed differences between BF and MRS running in a controlled setup,especially during the initial stance phase of running for the sagittal ankle and frontal rearfoot motion.Proposed barefoot features could be partly demonstrated w ith the Nike Free 3.0.Nevertheless, changes in design of the Nike Free 3.0 should be considered in order to m im ic BF movement even more closely.Foot strike at touch-down remained on the rearfoot,both in BF and MRS.

We believe a barefoot and individual running style and rearfoot strike pattern can be imitated by m inimal running shoes.Furtherdevelopmentof the Nike Free should focus on a rounded heel shape without any heel flare and a further reduction of the m idsole height.Consequently,minimal running shoes m ight serve as a training device to strengthen smallmuscles around the ankle jointas shown by Bru¨ggemann et al.21Future prospective studies are required to prove this beneficial aspect of m inimal running shoes and to investigate whether injury rates can eventually be reduced as shown by Potthast et al.22Finally,studies addressing the relationship of BF running and performance would be beneficial to address the contradicting results of recent studies.

Confl ict of interest

There are no confl icts of interest including financial,personalor other relationships w ith other people or organizations.

Acknow ledgm ent

The authors want to thank Nike Inc.for providing the m inimal running shoes for the current study.

1.Nigg B.Biomechanical considerations on barefoot movement and barefoot shoe concepts.Footwear Sci2009;1:73—9.

2.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.Nature2010;463:531—5.

3.Nigg B,Enders H.Barefoot running— some critical considerations.Footwear Sci2013;5:1—7.

4.Bonacci J,Saunders PU,Hicks A,Rantalainen T,Vicenzino BG, Spratford W.Running in a m inimalistand lightweightshoe is not the same as running barefoot:a biomechanical study.Br JSports Med2013;47:387—92.

5.Paquette MR,Zhang S,Baumgartner LD.Acute effects of barefoot, m inimal shoes and running shoes on lower limb mechanics in rear and forefoot strike runners.Footwear Sci2013;5:9—18.

6.Sinclair J,Greenhalgh A,Brooks D,Edmundson CJ,Hobbs SJ.The influence of barefoot and barefoot-inspired footwear on the kinetics and kinematics of running in comparison to conventional running shoes.Footwear Sci2013;5:45—53.

7.Squadrone R,GallozziC.Biomechanicaland physiologicalcomparison of barefoot and two shod conditions in experienced barefoot runners.J Sports Med Phys Fitness2009;49:6—13.

8.W illiams III DB,Green DH,Wurzinger B.Changes in lower extremity movement and power absorption during forefoot striking and barefoot running.Int J Sports Phys Ther2012;7:525—32.

9.Enders H,von Tscharner V,Nigg BM.The effects of preferred and nonpreferred running strike patterns on tissue vibration properties.J Sci Sport Med2014;17:218—22.

10.Kasmer ME,Liu XC,Roberts KG,Valadao JM.The relationship of foot strike pattern,shoe type,and performance in a 50-km trail race.J Strength Cond Res;2013.http://dx.doi.org/10.1519/JSC.0b013e3182a20ed4[Epub ahead of print].

11.Hatala KG,DingwallHL,Wunderlich RE,Richmond BG.Variation in foot strike patternsduring running among habitually barefootpopulations.PLoS One2013;8:e52548.http://dx.doi.org/10.1371/journal.pone.0052548.

12.Gruber AH,Silvernail JF,Brueggemann P,Rohr E,Ham ill J.Footfall patterns during barefoot running on harder and softer surfaces.Footwear Sci2013;5:39—44.

13.TenBroek TM,Rodrigues P,Frederick EC,Ham ill J.Effects of unknown footwear midsole thickness on running kinematics within the initial six m inutes of running.Footwear Sci2013;5:27—37.

14.Arnold JB,Bishop C.Quantifying foot kinematics inside athletic footwear:a review.Footwear Sci2013;5:55—62.

15.Wu G,Siegler S,Allard P,Kirtley C,LeardiniA,Rosenbaum D,etal.ISB recommendation on definitions of jointcoordinate system of various joints for the reporting of human joint motion—part I:ankle,hip,and spine.J Biomech2002;35:543—8.

16.So¨derkvist I,Wedin PA.Determ ining the movements of the skeleton using well-configured markers.J Biomech1993;26:1473—7.

17.Maiwald C,Sterzing T,Mayer T,M ilani T.Detecting foot-to-ground contact from kinematic data in running.Footwear Sci2009;1:111—8.

18.A ltman DG.Practical statistics for medical research.London:CRC Press;1991.

19.Bishop M,Fiolkowski P,Conrad B,Brunt D,Horodyski M.Athletic footwear,leg stiffness,and running kinematics.JAthl Train2006;41:387—92.

20.De Wit B,De Clercq D,Aerts P.Biomechanical analysis of the stance phase during barefoot and shod running.J Biomech2000;33:269—78.

21.Bru¨ggemann G-P,Goldmann J,Potthast W.Effects and evaluation of functional footwear.Proceedings Orthopaedie und Reha Technik,Leipzig, Germany,May 21—24,2008.

22.Potthast W,Braunstein B,Niehoff A,Bru¨ggemann G-P.The choice of training footwearhas an effecton changes in morphology and function of footand shank muscles.International Society of Biomechanics Congress, Beijing,China,August 22—27,2005.p.668—70.

Received 2 September 2013;revised 6 February 2014;accepted 17 March 2014

*Corresponding author.Department of Food and Nutrition,and Sport Science,University of Gothenburg,Go¨teborg 40530,Sweden.

E-mail address:stefan.grau@gu.se(S.Grau)

Peer review under responsibility of Shanghai University of Sport

2095-2546/$-see front matter CopyrightⒸ2014,Shanghai University of Sport.Production and hosting by Elsevier B.V.A ll rights reserved. http://dx.doi.org/10.1016/j.jshs.2014.03.002