This article was downloaded by: [Eveline Graf] On: 23 May 2013, At: 04:28 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Footwear Science Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tfws20 The effect of shoe torsional stiffness on lower extremity kinematics and biomechanical risk factors for patellofemoral pain syndrome during running Eveline S. Graf a & Darren Stefanyshyn a a Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Calgary, Alberta, Canada Published online: 08 Nov 2012. To cite this article: Eveline S. Graf & Darren Stefanyshyn (2012): The effect of shoe torsional stiffness on lower extremity kinematics and biomechanical risk factors for patellofemoral pain syndrome during running, Footwear Science, 4:3, 199-206 To link to this article: http://dx.doi.org/10.1080/19424280.2012.679703 PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Footwear Science Vol. 4, No. 3, September 2012, 199 206 The effect of shoe torsional stiffness on lower extremity kinematics and biomechanical risk factors for patellofemoral pain syndrome during running Eveline S. Graf* and Darren Stefanyshyn Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Calgary, Alberta, Canada (Received 20 December 2011; final version received 22 March 2012) Objective: The effect of footwear torsional stiffness on lower extremity biomechanics is not well known, although there are indications that it could affect rearfoot and ankle kinematics. These variables have previously been linked to the development of patellofemoral pain syndrome (PFPS) in runners. Therefore, the aim of this study was to compare the rearfoot and ankle frontal plane kinematics and knee abduction angular impulse between shoes with different torsional stiffness during running. Methods: Nineteen experienced runners performed heel-toe running at 3.7 m s 1 in three running shoes with different torsional stiffness. Using surface-mounted markers and a force plate, the kinematics and kinetics of the rearfoot, ankle and knee were measured. Torsion, rearfoot and ankle eversion, tibial rotation, knee abduction impulse, and vertical ground reaction force (GRF) peak were compared between footwear conditions using repeated-measures ANOVA. Results: The torsion angle was significantly different between shoes but none of the other variables showed a difference between conditions. Focusing only on the part of the stance phase with forefoot ground contact, significant differences were reported in the range of motion (ROM) of rearfoot and ankle eversion. The differences in torsional stiffness of the running shoes did not alter the rearfoot/ankle kinematics or the knee angular impulse, which are variables that have been described as risk factors for PFPS. Conclusions: During heel-toe running, the shoe torsional stiffness does not seem to have an effect on the injury risk for PFPS. However, there are indications that, for movements performed mainly on the forefoot, this shoe characteristic could have relevance. Keywords: Heel-toe running; foot torsion; torsional stiffness; patellofemoral pain syndrome; ankle kinematics 1. Introduction Patellofemoral pain syndrome (PFPS) is the most prevalent running injury, accounting for about 25% of all running-related injuries (Macintyre et al. 1991, Taunton et al. 2002). It is an injury common in the physically active population, but the exact aetiology is not well understood (Cheung et al. 2006). PFPS has been defined as pain in the retropatellar knee when no other pathologies are present (Messier and Pittala 1988, Witvrouw et al. 2000, Cowan et al. 2001). A study looking at the knee kinetics during running in a prospective design found that athletes who developed PFPS showed significantly higher knee abduction impulses during running at a controlled speed (Stefanyshyn et al. 2006). It was speculated that increased moments lead to higher contact forces in the joint. Increased contact forces combined with cumulative loading, which can be expressed using impulse variables, could lead to an overload in the patellofemoral joint. There are ambiguous results about the influence of rearfoot kinematics on PFPS development in running, but there are indications for a possible relationship. Using a cadaver model and computer simulations, it has been shown that rearfoot eversion results in internal tibial rotation (Hintermann et al. 1994) and excessive eversion can delay the tibial external rotation in the midstance phase, which could cause stress in the tibiofemoral and patellofemoral joints (Tiberio 1987). The relationship between rearfoot kinematics and PFPS during heel-toe running has not been studied in a prospective manner but, using a retrospective approach, Dierks et al. (2011) found a significant difference in rearfoot eversion between subjects suffering from PFPS and a control group. Other studies, however, failed to find significant differences in *Corresponding author. Email: esgraf@ucalgary.ca ISSN 1942 4280 print/issn 1942 4299 online ß 2012 Taylor & Francis http://dx.doi.org/10.1080/19424280.2012.679703 http://www.tandfonline.com
200 E.S. Graf and D. Stefanyshyn lower extremity kinematics between injured and uninjured groups (Messier et al. 1991, Levinger and Gilleard 2007). Torsion of the foot is defined as rotation between the forefoot and rearfoot in the frontal plane (Stacoff et al. 1989), which reduces the coupling of the movement between the two foot segments (Steindler 1973, Segesser et al. 1989). For barefoot heel-toe running the foot strikes the ground first with the lateral aspect of the heel, followed immediately by rearfoot eversion. Between 10 and 20 ms after heel strike, while the rearfoot is still everting, the forefoot touches the ground with the lateral side. The forefoot undergoes an eversion movement until it is flat on the ground. It is assumed that through the torsional movement of the forefoot relative to the rearfoot, the forefoot eversion does not cause an increased rearfoot eversion (Stacoff et al. 1989). Comparing barefoot and shod heel-toe running showed that shoes could decrease torsion significantly (Kaelin et al. 1989, Stacoff et al. 1989, Morio et al. 2009). The amount of rearfoot eversion has been shown to be significantly higher for shod running, providing evidence for the assumption that limited torsion increases rearfoot eversion (Kaelin et al. 1989, Stacoff et al. 1989). However, more recent studies failed to find differences in ankle frontal plane motion for barefoot and shod running (Stacoff et al. 2000, Eslami et al. 2007). Although running shoe torsion is an important design criterion, the effect of shoes with different torsional stiffness on the kinematics and kinetics of running has not yet been studied. Should increased footwear torsional stiffness in fact cause an increased rearfoot eversion, the tibial rotation could be larger when running in torsional stiff shoes. Based on previous work, this could increase the risk for PFPS. Another risk factor for PFPS is increased knee angular impulse. The effect of the torsional stiffness of the shoe during running on this variable has not been studied before. Therefore, the purpose of this study was to compare the effect of shoes with different torsional stiffness on rearfoot and ankle frontal plane kinematics and knee abduction angular impulse. 2. Methods Nineteen male subjects (age 28.2 8.7 years; height 173.4 4.4 cm; mass 70.5 8.4 kg), who ran a minimum of 15 km per week were recruited for this study. Each subject was free of any lower extremity injury and fit a shoe of size US9. The study was approved by the institution s ethics board and subjects gave informed written consent prior to data collection. On a 25-m Table 1. Shoe modifications and torsional stiffness. Flexible Medium Stiff Upper material Soft Soft Stiff Midsole hardness (Shore) 45C 55C 55C Forefoot inversion stiffness (Nm/ ) 0.18 0.26 0.47 Forefoot eversion stiffness (Nm/ ) 0.15 0.35 0.54 track, each subject performed 10 trials of heel-toe running at 3.7 m s 1 ( 5%) with each shoe condition. A running speed of 3.7 m s 1 has been reported as the self-selected running speed in recreational runners (Zamparo et al. 2001). The running speed was monitored using two photocells spaced 1.9 m apart. An Adidas Supernova Cushion 6 running shoe was modified to alter the torsional stiffness by using different upper materials and midsole hardnesses (Table 1). The shoe torsional stiffness was determined using a standard approach used in the footwear industry. A shoe last with a cutout in the midfoot area was inserted into the shoe. The gap between the forefoot and rearfoot part of the last was held constant using a beam around which the forefoot and rearfoot element could rotate freely. The heel part of the shoe was secured onto a fixed plate while the forefoot part was secured onto a plate that allowed movement around the long axis of the shoe. This forefoot plate was deflected to 10 of inversion and eversion while the applied torque was measured and used to quantify the shoe stiffness. Differences in bending stiffness between shoes, determined through standard testing, were minimal. Three retro-reflective markers were attached to each of the following segments of the right shoe and leg: forefoot, rearfoot and shank (Figure 1). With each shoe condition, a static, upright standing neutral trial was performed with the feet aligned visually with the laboratory coordinate system. During this trial, additional markers were attached on the shoe at the locations of the first and fifth metatarsal head, and on the skin at the locations of the lateral and medial malleoli and the lateral and medial epicondyles of the femur to define the metatarsophalangeal, ankle and knee joints, respectively. The marker trajectories were collected using a Motion Analysis system with eight high-speed digital cameras (Motion Analysis Corp., Santa Rosa, CA, USA) with a sampling rate of 240 Hz. Ground reaction forces (GRF) were recorded simultaneously using a Kistler force plate operating at 2400 Hz (Kistler AG, Winterthur, Switzerland). Before the analysis, data were filtered using a lowpass Butterworth filter (fourth-order) with a cut-off
Footwear Science 201 to when only the rearfoot or the entire foot is touching the ground. The data of all shoe conditions were compared with repeated-measures ANOVA (a ¼ 0.05) with Bonferroni post-hoc analysis where applicable (SPSS 11.5, IBM, Armonk, NY, USA). The average values were calculated based on the mean of each subject. Figure 1. Marker set (medial marker on the forefoot is hidden). frequency of 12 Hz for kinematic and 60 Hz for kinetic data. Torsion angles were calculated with a finite helical axis approach (Graf et al. 2012) using Matlab Version 7.5 (The MathWorks, Natick, MA, USA). All other kinematic (rearfoot and ankle eversion range of motion (ROM), maximal rearfoot and ankle eversion, tibial rotation) and kinetic variables (vertical GRF and knee abduction impulse based on an inverse dynamics approach) were calculated using KinTrak Version 7.0 (Motion Analysis Corp.). The knee abduction impulse was determined as the time integral of the moment curve. Rearfoot eversion was calculated relative to the laboratory system while ankle eversion referred to the relative angle in the frontal plane between rearfoot and shank. The tibial rotation was also calculated at the ankle joint and was therefore equal to ankle abduction (Tiberio 1987). Additionally, ankle and rearfoot frontal plane ROM was quantified during forefoot ground contact only. Forefoot ground contact was determined as the time when the centre of pressure (COP) was distal to the line formed by the first and fifth metatarsal head. It was assumed that when only the forefoot is in contact with the ground, footwear modifications in the midfoot area could have a larger effect on proximal joints compared 3. Results The flexible and medium shoes both showed an average ROM of torsion during heel-toe running that was significantly different from the stiff shoe; however, there was no difference between the medium and the flexible shoe (Figure 2, Table 2). None of the variables calculated for the rearfoot (maximal eversion and eversion ROM), ankle (maximal eversion, eversion ROM and tibial rotation) and knee (abduction angular impulse) (Figure 3 6), as well as the maximal vertical GRF, showed any significant differences between footwear conditions (Table 2). The rearfoot and ankle eversion ROM was also calculated when the COP was underneath the forefoot. The flexible shoe was not different from the medium shoe in both variables. For the stiff shoe, however, both movements were significantly smaller than for the other conditions (Table 3). 4. Discussion The purpose of this study was to quantify the effect of running shoes with different torsional stiffness on rearfoot and ankle kinematics and knee angular impulse. These variables have previously been related to the development of PFPS in runners. The footwear tested in this study led to different ROM for torsion, which is in agreement with early studies assessing the torsional movement of the foot during heel-toe running (Segesser et al. 1989, Stacoff et al. 1989). These early studies reported a change in rearfoot kinematics depending on the shoe stiffness. In the current study, however, the motion of the rearfoot relative to the laboratory condition or relative to the shank was not different between conditions when examining the entire stance phase. Therefore, the differences in torsion were the result of differences in only the forefoot frontal plane kinematics. A similar result was reported by Eslami et al. (2007), who measured foot and leg kinematics during running with a sandal and barefoot. Torsion was calculated for different sections of the stance phase. Even though none of the phases displayed a significant difference between barefoot and shod running, the mean torsion
202 E.S. Graf and D. Stefanyshyn Figure 2. Average torsion angle during the stance phase of heel-toe running (n ¼ 19). Table 2. Mean (SD) of all subjects of variables analysed over the whole stance phase. Flexible Medium Stiff p value Torsion ROM ( ) 6.5 (2.2) 6.0 (2.0) 4.7 (1.3) 50.001*y Rearfoot eversion ROM ( ) 10.8 (4.3) 10.6 (4.6) 10.3 (4.3) 0.20 Rearfoot maximal eversion ( ) 1.6 (3.5) 2.0 (3.3) 1.5 (3.9) 0.10 Ankle eversion ROM ( ) 12.6 (4.0) 12.4 (4.2) 12.5 (4.1) 0.88 Ankle maximal eversion ( ) 6.6 (3.2) 7.0 (3.2) 6.3 (4.1) 0.21 Maximal tibial rotation ( ) 8.6 (4.3) 9.0 (4.9) 8.4 (4.6) 0.18 Maximal vertical GRF (N) 1752 (206) 1753 (203) 1748 (216) 0.80 Knee abduction impulse (Nms) 8.3 (4.1) 8.1 (3.9) 7.9 (3.6) 0.42 Note: *Significant differences between the flexible and the stiff shoe, based on post-hoc analysis. ysignificant differences between the medium and the stiff shoe, based on post-hoc analysis. angle was up to 10 different, indicating some effect of footwear on the torsion ROM. A possible explanation for the different findings for rearfoot motion of the more recent studies compared to the studies by Segesser et al. (1989) and Stacoff et al. (1989) could be the measurement technique. Both early studies used only one camera filming the foot from behind. To minimize projection errors, the foot had to remain aligned with the camera orientation throughout the whole stance time (e.g. no adduction/abduction of the foot). This, however, is very difficult to achieve during an athletic movement such as running. It is therefore possible that a systematic change of the foot position relative to the camera depending on the footwear condition affected the measured rearfoot angle. The rearfoot and ankle kinematics were assessed in this study because previous work has indicated that excessive rearfoot eversion, which causes increased tibial rotation, could apply stress to the patellofemoral joint (Tiberio 1987). As no difference in rearfoot or ankle kinematics was found for the shoes assessed in this study, it appears that running shoe torsional stiffness does not alter the risk of PFPS caused by this mechanism. In general, the effect of shoe modifications on bone motion is limited. Previous studies assessing the rearfoot and shank motion using bone-anchored
Footwear Science 203 Figure 3. Average rearfoot eversion during the stance phase of heel-toe running (n ¼ 19). Figure 4. Average ankle eversion during the stance phase of heel-toe running (n ¼ 19). markers have shown that only very extreme modifications of the shoe heel geometry cause changes in rearfoot eversion and tibial rotation while smaller sole alterations do not have an effect on the bone kinematics (Stacoff et al. 2000, 2001). The alterations in the midfoot region of the shoe sole in the current study were not extreme enough to cause an effect on the rearfoot segment kinematics and it remains to
204 E.S. Graf and D. Stefanyshyn Figure 5. Average tibial rotation during the stance phase of heel-toe running (n ¼ 19). Figure 6. Average knee abduction moment during the stance phase of heel-toe running (n ¼ 19). be determined whether midfoot modifications can cause a change in the kinematics of proximal segments. The torsional stiffness of a shoe can be influenced by several factors including forefoot design, rearfoot design, specially designed torsion elements and even midsole and upper materials. In this investigation, different aspects of the footwear were used to influence its torsional stiffness. The flexible shoe had a softer
Footwear Science 205 Table 3. Mean (SD) of all subjects of variables analysed while centre of pressure under forefoot. Flexible Medium Stiff p value Rearfoot eversion ROM ( ) 9.8 (4.3) 9.5 (4.6) 8.3 (4.3) 50.001*y Ankle eversion ROM ( ) 10.6 (4.2) 10.1 (3.9) 8.6 (4.0) 50.001*y *Significant differences between the flexible and the stiff shoe, based on post-hoc analysis. ysignificant differences between the medium and the stiff shoe, based on post-hoc analysis. midsole material than the medium and the stiff shoes. The latter two shoes had different upper materials, which resulted in a torsional stiffer shoe for the stiff upper (Table 1). While those modifications did lead to differences in torsional stiffness, the effect on the midsole bending stiffness, determined through standardized testing, was minimal. Future studies are needed to determine whether different factors (midsole hardness, upper stiffness) have different effects on the kinetics and kinematics during running. A limitation of the current study was that the foot kinematics were estimated through markers placed on the footwear. Morio et al. (2009) compared running in sandals with a soft and a hard midsole stiffness and quantified the torsion based on markers placed on the foot as well as on the sole of the sandals. The comparison of torsion between the two sandals based on the skin markers revealed a small but significant difference in ROM. Measuring torsion based on markers placed on the sole of the sandals showed greater differences in ROM than when using skin markers. Even though Morio et al. (2009) used sandals, whereas in the current study running shoes were tested, the results are comparable. It has been shown that during running there is no difference in the measured eversion ROM between sandals and running shoes (Barnes et al. 2010). Therefore, the actual foot torsion might be smaller than the value determined from the shoe kinematics. The comparison of knee abduction impulse between footwear conditions revealed no differences. Because the rearfoot and ankle frontal plane kinematics as well as the maximal vertical GRF were not different between the tested shoes, this result is to be expected, based on mechanical considerations. There is only very limited knowledge about the influence of footwear on knee abduction impulse during repetitive motions such as walking and running. A recent study showed a decrease in knee abduction impulse during walking with lateral wedged shoes for subjects with osteoarthritis of the medial knee compartment (Hinman et al. 2012). That study showed that shoe modifications at the rearfoot could in fact alter the knee joint loading during a motion where the rearfoot is the contact point between foot and ground at the beginning of the stance phase. Torsion of the foot describes a movement between forefoot and rearfoot and the torsional stiffness of a shoe is assumed to modify the motion coupling between these two segments. It can therefore be argued that, when focusing on the effect of footwear torsional stiffness on the movement of proximal joints, the phase when the forefoot is the contact point between foot and ground is of more interest. In fact, the frontal plane ROM of the rearfoot (relative to the laboratory) and the ankle, calculated when the COP was underneath the forefoot, showed significant differences between shoe conditions, with larger movement with the more flexible shoes. This resulted primarily from large differences at the end of the stance phase, where the rearfoot or ankle was in a larger inversion with the flexible shoe compared to the other conditions (Figures 3 and 4). During barefoot running, at the end of the stance phase the ankle is at approximately 10 of inversion (Stacoff et al. 2000, Pohl et al. 2007). The final inversion reported in this study was between 2 and 4 (smaller angle for stiff shoe). The current study did not measure barefoot motion; therefore, a direct comparison with the previous studies is limited. However, it can be speculated that the footwear restricted the rearfoot motion. Therefore, torsional stiffness of footwear may have a significant influence on proximal joint kinetics and kinematics during movements with mainly forefoot ground contact (e.g. forefoot running, cutting movement); however, future studies are necessary to examine this relationship. 5. Conclusions This study examined the influence of footwear with different torsional stiffness on rearfoot and ankle kinematics as well as knee joint kinetics. Although the torsional movement was altered according to the shoe stiffness, when looking at the entire stance phase no effect on the kinematics of the ankle or the knee abduction impulse was found. This indicated that the shoe torsional stiffness, which is a modification in the midfoot region, does not alter the risk for the
206 E.S. Graf and D. Stefanyshyn development of PFPS according to previously proposed risk factors during heel-toe running. Acknowledgements The authors would like to thank adidas AG for providing the footwear used in this study. References Barnes, A., Wheat, J., and Milner, C.E., 2010. Use of gait sandals for measuring rearfoot and shank motion during running. Gait and Posture, 32 (1), 133 135. Cheung, R.T.H., Ng, G.Y.F., and Chen, B.F.C., 2006. Association of footwear with patellofemoral pain syndrome in runners. Sports Medicine, 36 (3), 199 205. Cowan, S.M., et al., 2001. Delayed onset of electromyographic activity of vastus medialis obliquus relative to vastus lateralis in subjects with patellofemoral pain syndrome. Archives of Physical Medicine and Rehabilitation, 82 (2), 183 189. Dierks, T., et al., 2011. Lower extremity kinematics in runners with patellofemoral pain during a prolonged run. Medicine and Science in Sports and Exercise, 43 (4), 693 700. Eslami, M., et al., 2007. Forefoot-rearfoot coupling patterns and tibial internal rotation during stance phase of barefoot versus shod running. Clinical Biomechanics, 22 (1), 74 80. Graf, E.S., Wright, I.C., and Stefanyshyn, D.J., 2012. Effect of relative marker movement on the calculation of the foot torsion axis using a combined Cardan angle and helical axis approach. Computational and Mathematical Methods in Medicine. Available from: http://www.hindawi.com/ journals/cmmm/aip [Accessed 27 February 2012]. Hinman, R.S., et al., 2012. Lateral wedge insoles for medial knee osteoarthritis: effects on lower limb frontal plane biomechanics. Clinical Biomechanics (Bristol, Avon), 27 (1), 27 33. Hintermann, B., et al., 1994. Transfer of movement between calcaneus and tibia in vitro. Clinical Biomechanics (Bristol, Avon), 9 (6), 349 355. Kaelin, X., et al., 1989. Torsion: an important factor of the running shoe design to reduce the risk of injuries (in German). Orthopaedische Praxis, 25 (7), 469 473. Levinger, P. and Gilleard, W., 2007. Tibia and rearfoot motion and ground reaction forces in subjects with patellofemoral pain syndrome during walking. Gait and Posture, 25 (1), 2 8. Macintyre, J.G., et al., 1991. Running injury: a clinical study of 4173 cases. Clinical Journal of Sport Medicine, 1 (2), 81 87. Messier, S.P. and Pittala, K., 1988. Etiologic factors associated with selected running injuries. Medicine and Science in Sports and Exercise, 20 (5), 501 505. Messier, S.P., et al., 1991. Etiologic factors associated with patellofemoral pain in runners. Medicine and Science in Sports and Exercise, 23 (9), 1008 1015. Morio, C., et al., 2009. The influence of footwear on foot motion during walking and running. Journal of Biomechanics, 42 (13), 2081 2088. Pohl, M.B., Messenger, N., and Buckley, J.G., 2007. Forefoot, rearfoot and shank coupling: effect of variations in speed and mode of gait. Gait and Posture, 25 (2), 295 302. Segesser, B., et al., 1989. Torsion a new concept in sport shoe construction (in German). Sportverletzung Sportschaden, 4, 167 182. Stacoff, A., et al., 1989. The torsion of the foot in running. International Journal of Sport Biomechanics, 5, 375 390. Stacoff, A., et al., 2000. Tibiocalcaneal kinematics of barefoot versus shod running. Journal of Biomechanics, 33 (11), 1387 1395. Stacoff, A., et al., 2001. Effects of shoe sole construction on skeletal motion during running. Medicine and Science in Sports and Exercise, 33 (2), 311 319. Stefanyshyn, D.J., et al., 2006. Knee angular impulse as a predictor of patellofemoral pain in runners. American Journal of Sports Medicine, 34 (11), 1844 1851. Steindler, A., 1973. Kinesiology of the human body under normal and pathological conditions, 4th ed. Springfield: Charles C. Thomas. Taunton, J.E., et al., 2002. A retrospective case-control analysis of 2002 running injuries. British Journal of Sports Medicine, 36, 95 101. Tiberio, D., 1987. The effect of excessive subtalar joint pronation on patellofemoral mechanics: a theoretical model. Journal of Orthopaedic and Sports Physical Therapy, 9 (4), 160 165. Witvrouw, E., et al., 2000. Intrinsic risk factors for the development of anterior knee pain in an athletic population: a two-year prospective study. American Journal of Sports Medicine, 28 (4), 480 489. Zamparo, P., et al., 2001. The self selected speed of running in recreational long distance runners. International Journal of Sports Medicine, 22 (8), 598 604.