- Email: [email protected]
Measuring medial longitudinal arch deformation during gait. A reliability study
Gait & Posture 35 (2012) 400–404
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Measuring medial longitudinal arch deformation during gait. A reliability study Jesper Bencke a,*, Ditte Christiansen b, Kathrine Jensen b, Anne Okholm b, Stig Sonne-Holm a, Thomas Bandholm a,c a
Gait Analysis Laboratory, Department of Orthopaedic Surgery, Copenhagen University Hospital, Hvidovre, Denmark University College Øresund, Faculty of Physiotherapy, Copenhagen, Denmark c Clinical Research Centre and Department of Physical Therapy, Copenhagen University Hospital, Hvidovre, Denmark b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 March 2011 Received in revised form 26 October 2011 Accepted 30 October 2011
Clinical evaluation of medial longitudinal arch deformation (MLAD) during walking gait is often estimated from static measures of e.g. navicular drop (ND) measured during quiet standing. The aim of the present study was to test the reliability of a new three-dimensional method of measuring the MLAD during gait and to compare this method with a static measure and a 2D dynamic method. Fifty-two feet (26 healthy male participants) were tested twice 4–9 days apart in a biomechanical gait analysis laboratory using a 3D three-marker foot model, a 2D video-based model for the measurement of MLAD during gait, and ND for measurements of MLAD during quiet standing. The 3D method showed the highest test-retest reliability among the measurements of MLAD. Furthermore, the ND showed only moderate correlation with both measurements of MLAD during gait. The new 3D method was found to be highly reliable and showed that ND obtained during quiet standing could not predict the MLAD during gait. The 3D method, or alternatively the 2D method, may be used in clinical settings as reliable methods for easy estimation of the foot longitudinal stability. ß 2011 Elsevier B.V. All rights reserved.
Keywords: Biomechanics Clinical gait analysis Foot Medial longitudinal arch Reliability
1. Introduction The complex anatomical structure of the foot allows many multi-planar movements within the foot. During loading of the foot, the greatest amount of motion occurs in the sagittal plane around the talonavicular joint [1,2], which can also be described as deformation of the medial longitudinal arch (MLA). The stability of the MLA may therefore be determined by the degree of MLA deformation (MLAD), and during gait the MLAD is dependent on both passive and active anatomical structures [3]. The plantar aponeurosis is shown to be the most important passive arch support during the stance phase [4], along with the spring ligament and the short and long plantar ligaments [5], while lower leg muscles and intrinsic foot muscles are important active supporters of the MLA, with the posterior tibialis muscle playing a dominant role [4,6]. Biomechanical evaluation of the MLAD during dynamic activity, such as gait, is not possible in daily clinical practice, and therefore a variety of static measures have prevailed as predictors of foot posture during dynamic conditions. Clinical measurements of navicular
* Corresponding author at: Gait Analysis Laboratory, sect. 247, Department of Orthopaedic Surgery, Hvidovre University Hospital, Kettegaard Alle 30, DK-2650 Hvidovre, Denmark. Tel.: +45 3632 6932; fax: +45 3632 3782. E-mail addresses: [email protected], [email protected] (J. Bencke). 0966-6362/$ – see front matter ß 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2011.10.360
height correlate closely to radiographic measurements of MLA [7], and the method known as the Navicular Drop Test (NDT), first described by Brody [8], is one of the most extensively used tests for description of foot posture, although reliability may be debatable. The results of intra-tester and inter-tester reliability investigations of the NDT range from poor to excellent, with an Intraclass Correlation Coefficient (ICC) from 0.33 [9] to 0.97 [10–12]. The great variance in reliability among studies has been suggested to arise from the difficulty in consistently placing the subtalar joint in its neutral position and difficulty in palpating the navicular tubercle [13]. Despite the variance in reliability, the NDT has been considered the most reliable static test for prediction of MLA during dynamic activity [13]. The NDT has been used in attempts to clinically distinguish subjects with pathological conditions from healthy. Some sports-related injuries with either a high incidence or severe consequences for the athlete have been related to MLA stability. Medial Tibial Stress Syndrome (MTSS) is the most common injury in running accounting for nearly 60% of all overuse leg injuries [14]. Measurements of the navicular drop and other static measures of foot pronation have in some studies been related to incidence of MTSS [10,15], while other studies found no relation between navicular drop and MTSS [16]. Injury to the anterior cruciate ligament (ACL) is one of the most invalidating lower extremity injuries with incidences of 0.31 pr 1000 h of play [17]. The potential of the NDT to discriminate between subjects with pathological conditions is ambiguous as shown in retrospective studies of injuries of the ACL [11,18]
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
401
Controversy also exists as to whether patello-femoral pain syndrome, representing about 25% of sports-related knee injuries [19], is related to MLA stability [20,21]. On the other hand, studies using dynamic measures of MLAD have been able to distinguish subjects with lower leg overuse injuries [15]. The determination of dynamic MLAD predicted from static measurements has only been scarcely described [22,23], and MLAD during gait may not necessarily be described by static measures of foot posture [12,24]. Overloading injuries in the lower leg and foot occur as a result of dynamic activities, obviously due to the greater loading of the supporting structures. Therefore, it may be reasonable to apply measurements of MLAD during dynamic conditions, as these may be more valid as predictors of injury risk. The main objective of the present study was to determine the validity and reliability of two clinical measurements of MLAD during dynamic loading (gait) previously published [15]. A secondary objective was to determine the validity and reliability of a commonly used static clinical measurement of MLAD (the navicular drop method) and its relation to MLAD during gait. It was hypothesized that the reliability of the dynamic MLAD measurements would be very high, and that the navicular drop method would not be highly related to dynamic MLAD. 2. Patients and methods The study was designed as an intra-tester reliability study with 4–9 days between test and retest. The intra-tester reliability was expected to be at least 0.7 and hopefully 0.9, and it was found that a sample size of more than 19 subjects would be adequate for obtaining a statistical power of 80%. Twenty-six male participants aged 20–51 years, with no lower leg or foot pathologies and no lower extremity orthopaedic injury within the last 6 months, volunteered as participants in the study. Both feet of each subject were studied, making a total of 52 ft. No females were included to avoid possible changes in ligamentous laxity due to hormonal fluctuations. The participants were informed of the methods used and their rights as participants according to the Helsinki Declaration, and written informed consent were obtained from all participants. 3. Measurements One static and two dynamic measurements were made of foot MLA.
Fig. 1. A subject placed in tandem stance with each foot on a forceplate, and the leg of interest positioned with a vertical tibia.
3.2. Dynamic tests For the dynamic measurements, three anatomical points were chosen: The medial aspect of the head of the first metatarsal bone (MH1), the medial prominence of the navicular tuberosity and on the medial side of the calcaneus 4.0 cm from the most posterior point and 3.5 cm from the ground. The MH1 and the navicular tuberosity were palpated and the exact position of the heel marker was determined by placing one side of a custom-built device between a wall and the posterior point of the heel and marking the position with a marker pen through the hole in the device on the medial side of the heel (see Fig. 2). For the 2D dynamic measurements, the three points were marked with a marker pen, and the angle between these marks was defined as the MLA [15].
3.1. Static test The NDT was modified from Brody’s test [8]. In the present study, the participants stood in tandem stance with one leg in front and each leg on a force plate for registration of weight distribution (Fig. 1). The investigator manipulated the subtalar joint into neutral position, marked the navicular tuberosity, and measured the height of the navicular tuberosity over the floor with a caliper in standing with a vertical tibia and approximately 20% of bodyweight on the foot. The neutral position of the subtalar joint was determined by palpating the head of the talus both laterally and medially directly inferior and anterior to the malleoles, and after having the subject performing unloaded eversions and inversions, the neutral position was established [8]. The subject was subsequently allowed to stand normally with a vertical tibia and approximately 80% of bodyweight on the foot, and the height of the navicular tuberosity was measured. The modified ND test procedure was repeated three times for each leg, and the average difference expressed the ND in mm.
Fig. 2. The custom-built device used for exact positioning of the heel marker on the medial side of the calcaneus 4.0 cm from the most posterior point and 3.5 cm from the ground. A pen was inserted through a little hole in the device, while the subject was standing with the device between the heel and wall.
402
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
The participants walked at a natural, self-selected pace along a 6 m path with a pressure mapping mat (Emed, Novel, Germany) mounted in the floor, and a standard digital video camera placed midway on the path filmed perpendicular to the walking direction with a frame rate of 50 Hz. This frame rate is the normal frame rate in custom video cameras. Previous studies have used video analyses with similar frame rates with adequate results for gait in humans [25]. The ground reaction forces were recorded in synchronization with the video recordings. Specific characteristic instants of the vertical ground reaction force curve were used to determine five points within the stance phase of gait that were, in turn, used to determine the MLA angles from the corresponding video frames. The angle of the MLA was determined at (1) initial heel strike with no load on the MLA, (2) at the instant of first force peak, (3) at midstance, defined as the instant of lowest force between the two force peaks, (4) at the instant of peak ground reaction push-off force, when the heel has been lifted from the ground and the MLA was loaded and (5) at toe-off (see Fig. 3). For the 3D kinematic measurements three markers were placed on the marked positions as described above. The 3D angle was modified to be the angle between these three markers in a vertical plane, which incorporates the heel and MH1 marker. This modified 3D angle represented the MLA during walking. The 3D coordinates of the markers were recorded at 100 Hz by an 8-camera VICON 612 system and processed with Vicon workstation (vers. 4.6, Oxford Metrics, Oxford, UK), and angles were calculated using a customwritten Matlab macro (vers.7.2, MatWorks Inc., Natick, Massachusetts, U.S.A.). The participants walked along a 10 m path with 2 force plates mounted in the floor, and the ground reaction forces were recorded in synchronization with the 3D marker recordings. As in the 2D measurements, the angles of the MLA were calculated
at specific instants in stance phase determined by vertical force parameters. In both measurements, the average values from 5 trials for each foot were used for the statistical analyses. Based on preliminary examination of the MLA measurements, the MLAD was defined as the difference between angle 4, representing the angle of most dynamic loading and angle 1, representing the most unloaded and thus the most neutral angle (see Fig. 3) in both 2D and 3D dynamic measurements. 3.3. Statistical analyses As all dependent variables were normally distributed, parametric tests were applied. Reliability of MLAD, angles 1 and 4 from both 2D and 3D dynamic measurements, and the ND were assessed using ICC type 2.1, and standard error of measurement (SEM). pffiffiffiStandard error of measurement was calculated as SEM ¼ SDdiff = 2; where SDdiff is the standard deviation of the difference in scores between the test and retest values. Differences between test results of the 2D and 3D test were analyzed using a paired Student’s t-tests. The relationship between static and dynamic measurements was assessed using Pearson’s correlation coefficients. Level of significance was set at p < 0.05. 4. Results The results of the first test session revealed that the 2D angles were significantly less than the 3D angles (mean (SD), angle 1: 150.3 (10.8) vs. 154.9 (10.2), respectively, p < 0.001), and angle 4: 159.8 (12.6) vs. 163.1 (11.5), respectively, p < 0.001) (Table 1), and there was a significant difference (p < 0.001) between 2D and 3D MLAD in that a systematically larger MLAD was found in the 2D
Fig. 3. A representative trial showing the deformation of the medial longitudinal arch (MLA) in degrees in relation to the vertical groundreaction force (GRF) in Newton. The circles indicate the different instants of the foot ground contact period where the MLA is calculated and the photos below illustrate the position of the foot at the specified time of the contact period.
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
403
Table 1 Descriptive data and reliability results. Parameter 3D 3D 2D 2D *
angle angle angle angle
1 4 1 4
(8) (8) (8) (8)
Test mean (SD)
Retest mean (SD)
Test–retest difference (SD)
ICC (2.1)
95% conf. Int.
SEM
154.9 163.1 150.3 159.8
154.9 163.0 150.5 160.0
0.0 0.1 0.2 0.2
0.95 0.95 0.95 0.95
0.92–0.97 0.92–0.97 0.91–0.97 0.92–0.97
3.4 3.5 3.4 3.8
(10.2) (11.5) (10.8)* (12.6)*
(10.5) (11.9) (11.4)* (13.0)*
(4.8) (5.0) (4.8) (5.4)
Denotes significant difference (p < 0.001) from respective 3D parameter.
Table 2 Descriptive data and reliability results of MLAD and ND. Parameter
Test mean (SD)
Retest mean (SD)
Test–retest difference (SD)
ICC (2.1)
95% conf. Int.
SEM
3D MLAD (8) 2D MLAD (8) ND (mm)
8.3 (2.8) 9.6 (3.4)* 7.2 (3.6)
8.1 (2.6) 9.5 (3.3)* 7.0 (3.9)
0.2 (1.1) 0.1 (1.9) 0.1 (2.4)
0.95 0.91 0.88
0.92–0.97 0.84–0.95 0.80–0.93
0.8 1.4 1.7
*
Denotes significant difference (p < 0.001) from 3D MLAD.
Table 3 Relationship between variables. Parameter
3D MLAD (8)
2D MLAD (8)
ND (mm)
3D MLAD (8)
–
2D MLAD (8)
–
0.76 p < 0.001 –
0.56 p < 0.001 0.61 p < 0.001
dynamic measurements (Table 2). The reliability measurements of static and dynamic deformation showed higher reliability in the dynamic tests than the static ND test (Table 2). The dynamic 3D MLAD showed less correlation with the static ND test (r = 0.56, p < 0.001) than with the 2D MLAD (r = 0.76, p < 0.001) (see Table 3). 5. Discussion In daily clinical examinations, time for laboratory testing is limited and demands for simple, reliable, and valid methods of testing are increasing. The present study investigated the reliability of sagittal plane variables of MLA during gait, using a simple modified 3D kinematic method. Three-dimensional studies have investigated the MLA either as the angle between three medially placed markers [2,26], as in the present study, or as angles between forefoot and hindfoot segments [27–29]. The range of MLA deformation in these studies ranges from about 58 [27] to about 128 [29]. The results on MLAD from this study (8.18 2.88) therefore concur with other studies, although the absolute values of MLA do not match the mentioned studies due to differences in marker placements. The reliability of 3D MLAD was excellent and comparable to the reliability measures reported in other 3D MLA studies [2,27]. The present method involved only three markers on each foot, and the total test time was less than 20 min. This method may therefore be more time-saving than other, more advanced, 3D analyses, which in some models require application of 17 or 19 markers [28,29]. The relative simplicity of the present model, compared to the advanced models, may increase its potential for clinical use. Besides reliability, the dynamic 3D MLAD test reflects how the supporting structures are responding to the external loads during dynamic activity. Since overloading injuries result from excess dynamic loading, dynamic tests like the 3D MLAD may be a better clinical predictor of overloading injury risk than static tests. The clinical significance of this MLAD method was evident in a study of MLAD differences between athletes with Medial Tibial Stress Syndrome and matched healthy athletes using the same procedures
as in the present study [15]. Even though other studies have shown good correlations between measures of MLA during gait midstance and static measures of MLA [22,23], the results of the present study show only a moderate correlation (Pearson’s, r = 0.56, p < 0.001) between the static ND test and the dynamic MLAD (Table 3). It could be speculated, that this lack of correlation may in part be explained by a higher loading of the MLA in the push-off situation compared to quiet standing. It has been shown, that the center of pressure is moving forward during the contact phase of walking, reaching a position under the forefoot during the push-off phase, and this increased external moment arm contributes greatly to the peak plantarflexor moment seen during the terminal stance phase in walking [30]. This is in contrast to quiet standing, where the center of pressure is located under the midfoot with much smaller plantarflexor moments. Assuming the foot as a two-segment structure in the sagittal plane with a hinge joint placed at the navicular bone between the forefoot and the hindfoot, the increased plantarflexor moment during push-off would correlate to an external moment attempting to dorsiflex the forefoot with respect to the hindfoot. The lack of good correlation between static and dynamic measures could thus be a result of this extra load being counteracted differently between participants. This would further imply that a static test measuring the ND might only have limited validity as predictor of MLAD during gait. Future investigations using kinetic foot models may assess the magnitude of these moments. In an attempt to make the method applicable in daily clinical practice, the 3D measurements were compared to the 2D measurements obtained by a single digital video camera placed perpendicularly to the direction of walking. There was a significant systematic difference between MLAD results from the two methods, which may be due to variations in foot progression angles observed in normal subjects. The repeatability of the 2D angles during gait was high and matched the 3D ICCs and SEMs. The correlation between the 3D measurement and the 2D measurement was higher than for the 3D measurement and the static ND test, but not excellent. This may be due to the slight variations of foot progression angles that are likely to occur during normal walking. However, the video based 2D measurements may be better than the static ND test in assessing the MLAD developed during gait, but may still have limited correlation to the probably more valid 3D MLAD, as the correlation coefficient (r = 0.76) suggests that 2D video measurements explain less than 60% of the variation of the 3D MLAD. Still, the 2D MLAD proves highly reproducible and may therefore be used in intervention studies, for example. Furthermore, this method may easily be applied in clinical settings without a biomechanical motion analysis system.
404
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
The major limitation of the presented new method is that it only measures the longitudinal medial arch, although the MLAD may be a composite of deformations in all three planes. Therefore, e.g. excess pronation will not be measured isolated, and the clinical interpretation may therefore be performed with regard to this limitation. The dynamic methods examined furthermore demand access to a camera based movement analysis system with a forceplate or similar, which may be a practical limitation. In conclusion, the new and simple 3D method of measuring deformation of the MLA during gait showed excellent reliability, and was more reliable than a dynamic 2D method and a traditional static method, as was hypothezised. Furthermore, the results of the present study show that static measurements of MLAD such as the ND may have limited validity as predictors of MLAD during gait. Being relatively quick to perform in a biomechanical laboratory setting, the 3D method may be used in clinical examinations of unstable feet scheduled for surgery or as an outcome measure following surgery or rehabilitation. Further studies may evaluate the clinical validity of the presented method in terms of ability to detect clinically relevant intervention improvements or predicting risk of overuse injuries like medial tibial stress syndrome or plantar fasciitis. Acknowledgements The authors like to thank Derek Curtis, PT, MSc, for his valuable assistance in reading through the final version of the manuscript and correcting misspellings and grammatical errors. Conflict of interest There are no duplicate publication elsewhere of any part of the work, and there are no commercial interests in the results of the present study, or other conflicts of interest. References [1] Winson I, Lundberg A, Bylund C. The pattern of motion of the longitudinal arch of the foot. The Foot 1994;4:151–4. [2] Hunt AE, Smith RM, Torode M, Keenan AM. Inter-segment foot motion and ground reaction forces over the stance phase of walking. Clin Biomech (Bristol Avon) 2001;16(7):592–600. [3] Richie Jr DH. Biomechanics and clinical analysis of the adult acquired flatfoot. Clin Podiatr Med Surg 2007;24(4):617–44. vii.. [4] Thordarson DB, Schmotzer H, Chon J, Peters J. Dynamic support of the human longitudinal arch. A biomechanical evaluation. Clin Orthop Relat Res 1995;316:165–72. [5] Tao K, Ji WT, Wang DM, Wang CT, Wang X. Relative contributions of plantar fascia and ligaments on the arch static stability: a finite element study. Biomed Tech (Berl) 2010;55(5):265–71. [6] Fiolkowski P, Brunt D, Bishop M, Woo R, Horodyski M. Intrinsic pedal musculature support of the medial longitudinal arch: an electromyography study. J Foot Ankle Surg 2003;42(6):327–33.
[7] Saltzman CL, Nawoczenski DA, Talbot KD. Measurement of the medial longitudinal arch. Arch Phys Med Rehabil 1995;76(1):45–9. [8] Brody DM. Techniques in the evaluation and treatment of the injured runner. Orthop Clin North Am 1982;13(3):541–58. [9] Vinicombe A, Raspovic A, Menz HB. Reliability of navicular displacement measurement as a clinical indicator of foot posture. J Am Podiatr Med Assoc 2001;91(5):262–8. [10] Bennett JE, Reinking MF, Pluemer B, Pentel A, Seaton M, Killian C. Factors contributing to the development of medial tibial stress syndrome in high school runners. J Orthop Sports Phys Ther 2001;31(9):504–10. [11] Allen MK, Glasoe WM. Metrecom measurement of navicular drop in subjects with anterior cruciate ligament injury. J Athl Train 2000;35(4):403–6. [12] Nielsen RG, Rathleff MS, Moelgaard CM, Simonsen O, Kaalund S, Olesen CG, et al. Video based analysis of dynamic midfoot function and its relationship with Foot Posture Index scores. Gait Posture 2010;31(1):126–30. [13] Menz HB. Alternative techniques for the clinical assessment of foot pronation. J Am Podiatr Med Assoc 1998;88(3):119–29. [14] Couture CJ, Karlson KA. Tibial stress injuries: decisive diagnosis and treatment of ‘shin splints’. Phys Sportsmed 2002;30(6):29–36. [15] Bandholm T, Boysen L, Haugaard S, Zebis MK, Bencke J. Foot medial longitudinal-arch deformation during quiet standing and gait in subjects with medial tibial stress syndrome. J Foot Ankle Surg 2008;47(2):89–95. [16] Hubbard TJ, Carpenter EM, Cordova ML. Contributing factors to medial tibial stress syndrome: a prospective investigation. Med Sci Sports Exerc 2009;41 (3):490–6. [17] Myklebust G, Maehlum S, Holm I, Bahr R. A prospective cohort study of anterior cruciate ligament injuries in elite Norwegian team handball. Scand J Med Sci Sports 1998;8(3):149–53. [18] Jenkins WL, Killian CB, Williams III DS, Loudon J, Raedeke SG. Anterior cruciate ligament injury in female and male athletes: the relationship between foot structure and injury. J Am Podiatr Med Assoc 2007;97(5):371–6. [19] Devereaux MD, Lachmann SM. Patello-femoral arthralgia in athletes attending a sports injury clinic. Br J Sports Med 1984;18(1):18–21. [20] Hetsroni I, Finestone A, Milgrom C, Sira DB, Nyska M, Radeva-Petrova D, et al. A prospective biomechanical study of the association between foot pronation and the incidence of anterior knee pain among military recruits. J Bone Joint Surg Br 2006;88(7):905–8. [21] Duffey MJ, Martin DF, Cannon DW, Craven T, Messier SP. Etiologic factors associated with anterior knee pain in distance runners. Med Sci Sports Exerc 2000;32(11):1825–32. [22] McPoil TG, Cornwall MW. Use of the longitudinal arch angle to predict dynamic foot posture in walking. J Am Podiatr Med Assoc 2005;95(2): 114–20. [23] Franettovich MM, McPoil TG, Russell T, Skardoon G, Vicenzino B. The ability to predict dynamic foot posture from static measurements. J Am Podiatr Med Assoc 2007;97(2):115–20. [24] Cashmere T, Smith R, Hunt A. Medial longitudinal arch of the foot: stationary versus walking measures. Foot Ankle Int 1999;20(2):112–8. [25] Alkjaer T, Simonsen EB, Dyhre-Poulsen P. Comparison of inverse dynamics calculated by two- and three-dimensional models during walking. Gait Posture 2001;13(2):73–7. [26] Tome J, Nawoczenski DA, Flemister A, Houck J. Comparison of foot kinematics between subjects with posterior tibialis tendon dysfunction and healthy controls. J Orthop Sports Phys Ther 2006;36(9):635–44. [27] Nakamura H, Kakurai S. Relationship between the medial longitudinal arch movement and the pattern of rearfoot motion during the stance phase of walking. J Phys Ther Sci 2003;15(1):13–8. [28] MacWilliams BA, Cowley M, Nicholson DE. Foot kinematics and kinetics during adolescent gait. Gait Posture 2003;17(3):214–24. [29] Stebbins J, Harrington M, Thompson N, Zavatsky A, Theologis T. Repeatability of a model for measuring multi-segment foot kinematics in children. Gait Posture 2006;23(4):401–10. [30] Haim A, Rozen N, Wolf A. The influence of sagittal center of pressure offset on gait kinematics and kinetics. J Biomech 2010;43(5):969–77.
Contents lists available at SciVerse ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Measuring medial longitudinal arch deformation during gait. A reliability study Jesper Bencke a,*, Ditte Christiansen b, Kathrine Jensen b, Anne Okholm b, Stig Sonne-Holm a, Thomas Bandholm a,c a
Gait Analysis Laboratory, Department of Orthopaedic Surgery, Copenhagen University Hospital, Hvidovre, Denmark University College Øresund, Faculty of Physiotherapy, Copenhagen, Denmark c Clinical Research Centre and Department of Physical Therapy, Copenhagen University Hospital, Hvidovre, Denmark b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 23 March 2011 Received in revised form 26 October 2011 Accepted 30 October 2011
Clinical evaluation of medial longitudinal arch deformation (MLAD) during walking gait is often estimated from static measures of e.g. navicular drop (ND) measured during quiet standing. The aim of the present study was to test the reliability of a new three-dimensional method of measuring the MLAD during gait and to compare this method with a static measure and a 2D dynamic method. Fifty-two feet (26 healthy male participants) were tested twice 4–9 days apart in a biomechanical gait analysis laboratory using a 3D three-marker foot model, a 2D video-based model for the measurement of MLAD during gait, and ND for measurements of MLAD during quiet standing. The 3D method showed the highest test-retest reliability among the measurements of MLAD. Furthermore, the ND showed only moderate correlation with both measurements of MLAD during gait. The new 3D method was found to be highly reliable and showed that ND obtained during quiet standing could not predict the MLAD during gait. The 3D method, or alternatively the 2D method, may be used in clinical settings as reliable methods for easy estimation of the foot longitudinal stability. ß 2011 Elsevier B.V. All rights reserved.
Keywords: Biomechanics Clinical gait analysis Foot Medial longitudinal arch Reliability
1. Introduction The complex anatomical structure of the foot allows many multi-planar movements within the foot. During loading of the foot, the greatest amount of motion occurs in the sagittal plane around the talonavicular joint [1,2], which can also be described as deformation of the medial longitudinal arch (MLA). The stability of the MLA may therefore be determined by the degree of MLA deformation (MLAD), and during gait the MLAD is dependent on both passive and active anatomical structures [3]. The plantar aponeurosis is shown to be the most important passive arch support during the stance phase [4], along with the spring ligament and the short and long plantar ligaments [5], while lower leg muscles and intrinsic foot muscles are important active supporters of the MLA, with the posterior tibialis muscle playing a dominant role [4,6]. Biomechanical evaluation of the MLAD during dynamic activity, such as gait, is not possible in daily clinical practice, and therefore a variety of static measures have prevailed as predictors of foot posture during dynamic conditions. Clinical measurements of navicular
* Corresponding author at: Gait Analysis Laboratory, sect. 247, Department of Orthopaedic Surgery, Hvidovre University Hospital, Kettegaard Alle 30, DK-2650 Hvidovre, Denmark. Tel.: +45 3632 6932; fax: +45 3632 3782. E-mail addresses: [email protected], [email protected] (J. Bencke). 0966-6362/$ – see front matter ß 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2011.10.360
height correlate closely to radiographic measurements of MLA [7], and the method known as the Navicular Drop Test (NDT), first described by Brody [8], is one of the most extensively used tests for description of foot posture, although reliability may be debatable. The results of intra-tester and inter-tester reliability investigations of the NDT range from poor to excellent, with an Intraclass Correlation Coefficient (ICC) from 0.33 [9] to 0.97 [10–12]. The great variance in reliability among studies has been suggested to arise from the difficulty in consistently placing the subtalar joint in its neutral position and difficulty in palpating the navicular tubercle [13]. Despite the variance in reliability, the NDT has been considered the most reliable static test for prediction of MLA during dynamic activity [13]. The NDT has been used in attempts to clinically distinguish subjects with pathological conditions from healthy. Some sports-related injuries with either a high incidence or severe consequences for the athlete have been related to MLA stability. Medial Tibial Stress Syndrome (MTSS) is the most common injury in running accounting for nearly 60% of all overuse leg injuries [14]. Measurements of the navicular drop and other static measures of foot pronation have in some studies been related to incidence of MTSS [10,15], while other studies found no relation between navicular drop and MTSS [16]. Injury to the anterior cruciate ligament (ACL) is one of the most invalidating lower extremity injuries with incidences of 0.31 pr 1000 h of play [17]. The potential of the NDT to discriminate between subjects with pathological conditions is ambiguous as shown in retrospective studies of injuries of the ACL [11,18]
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
401
Controversy also exists as to whether patello-femoral pain syndrome, representing about 25% of sports-related knee injuries [19], is related to MLA stability [20,21]. On the other hand, studies using dynamic measures of MLAD have been able to distinguish subjects with lower leg overuse injuries [15]. The determination of dynamic MLAD predicted from static measurements has only been scarcely described [22,23], and MLAD during gait may not necessarily be described by static measures of foot posture [12,24]. Overloading injuries in the lower leg and foot occur as a result of dynamic activities, obviously due to the greater loading of the supporting structures. Therefore, it may be reasonable to apply measurements of MLAD during dynamic conditions, as these may be more valid as predictors of injury risk. The main objective of the present study was to determine the validity and reliability of two clinical measurements of MLAD during dynamic loading (gait) previously published [15]. A secondary objective was to determine the validity and reliability of a commonly used static clinical measurement of MLAD (the navicular drop method) and its relation to MLAD during gait. It was hypothesized that the reliability of the dynamic MLAD measurements would be very high, and that the navicular drop method would not be highly related to dynamic MLAD. 2. Patients and methods The study was designed as an intra-tester reliability study with 4–9 days between test and retest. The intra-tester reliability was expected to be at least 0.7 and hopefully 0.9, and it was found that a sample size of more than 19 subjects would be adequate for obtaining a statistical power of 80%. Twenty-six male participants aged 20–51 years, with no lower leg or foot pathologies and no lower extremity orthopaedic injury within the last 6 months, volunteered as participants in the study. Both feet of each subject were studied, making a total of 52 ft. No females were included to avoid possible changes in ligamentous laxity due to hormonal fluctuations. The participants were informed of the methods used and their rights as participants according to the Helsinki Declaration, and written informed consent were obtained from all participants. 3. Measurements One static and two dynamic measurements were made of foot MLA.
Fig. 1. A subject placed in tandem stance with each foot on a forceplate, and the leg of interest positioned with a vertical tibia.
3.2. Dynamic tests For the dynamic measurements, three anatomical points were chosen: The medial aspect of the head of the first metatarsal bone (MH1), the medial prominence of the navicular tuberosity and on the medial side of the calcaneus 4.0 cm from the most posterior point and 3.5 cm from the ground. The MH1 and the navicular tuberosity were palpated and the exact position of the heel marker was determined by placing one side of a custom-built device between a wall and the posterior point of the heel and marking the position with a marker pen through the hole in the device on the medial side of the heel (see Fig. 2). For the 2D dynamic measurements, the three points were marked with a marker pen, and the angle between these marks was defined as the MLA [15].
3.1. Static test The NDT was modified from Brody’s test [8]. In the present study, the participants stood in tandem stance with one leg in front and each leg on a force plate for registration of weight distribution (Fig. 1). The investigator manipulated the subtalar joint into neutral position, marked the navicular tuberosity, and measured the height of the navicular tuberosity over the floor with a caliper in standing with a vertical tibia and approximately 20% of bodyweight on the foot. The neutral position of the subtalar joint was determined by palpating the head of the talus both laterally and medially directly inferior and anterior to the malleoles, and after having the subject performing unloaded eversions and inversions, the neutral position was established [8]. The subject was subsequently allowed to stand normally with a vertical tibia and approximately 80% of bodyweight on the foot, and the height of the navicular tuberosity was measured. The modified ND test procedure was repeated three times for each leg, and the average difference expressed the ND in mm.
Fig. 2. The custom-built device used for exact positioning of the heel marker on the medial side of the calcaneus 4.0 cm from the most posterior point and 3.5 cm from the ground. A pen was inserted through a little hole in the device, while the subject was standing with the device between the heel and wall.
402
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
The participants walked at a natural, self-selected pace along a 6 m path with a pressure mapping mat (Emed, Novel, Germany) mounted in the floor, and a standard digital video camera placed midway on the path filmed perpendicular to the walking direction with a frame rate of 50 Hz. This frame rate is the normal frame rate in custom video cameras. Previous studies have used video analyses with similar frame rates with adequate results for gait in humans [25]. The ground reaction forces were recorded in synchronization with the video recordings. Specific characteristic instants of the vertical ground reaction force curve were used to determine five points within the stance phase of gait that were, in turn, used to determine the MLA angles from the corresponding video frames. The angle of the MLA was determined at (1) initial heel strike with no load on the MLA, (2) at the instant of first force peak, (3) at midstance, defined as the instant of lowest force between the two force peaks, (4) at the instant of peak ground reaction push-off force, when the heel has been lifted from the ground and the MLA was loaded and (5) at toe-off (see Fig. 3). For the 3D kinematic measurements three markers were placed on the marked positions as described above. The 3D angle was modified to be the angle between these three markers in a vertical plane, which incorporates the heel and MH1 marker. This modified 3D angle represented the MLA during walking. The 3D coordinates of the markers were recorded at 100 Hz by an 8-camera VICON 612 system and processed with Vicon workstation (vers. 4.6, Oxford Metrics, Oxford, UK), and angles were calculated using a customwritten Matlab macro (vers.7.2, MatWorks Inc., Natick, Massachusetts, U.S.A.). The participants walked along a 10 m path with 2 force plates mounted in the floor, and the ground reaction forces were recorded in synchronization with the 3D marker recordings. As in the 2D measurements, the angles of the MLA were calculated
at specific instants in stance phase determined by vertical force parameters. In both measurements, the average values from 5 trials for each foot were used for the statistical analyses. Based on preliminary examination of the MLA measurements, the MLAD was defined as the difference between angle 4, representing the angle of most dynamic loading and angle 1, representing the most unloaded and thus the most neutral angle (see Fig. 3) in both 2D and 3D dynamic measurements. 3.3. Statistical analyses As all dependent variables were normally distributed, parametric tests were applied. Reliability of MLAD, angles 1 and 4 from both 2D and 3D dynamic measurements, and the ND were assessed using ICC type 2.1, and standard error of measurement (SEM). pffiffiffiStandard error of measurement was calculated as SEM ¼ SDdiff = 2; where SDdiff is the standard deviation of the difference in scores between the test and retest values. Differences between test results of the 2D and 3D test were analyzed using a paired Student’s t-tests. The relationship between static and dynamic measurements was assessed using Pearson’s correlation coefficients. Level of significance was set at p < 0.05. 4. Results The results of the first test session revealed that the 2D angles were significantly less than the 3D angles (mean (SD), angle 1: 150.3 (10.8) vs. 154.9 (10.2), respectively, p < 0.001), and angle 4: 159.8 (12.6) vs. 163.1 (11.5), respectively, p < 0.001) (Table 1), and there was a significant difference (p < 0.001) between 2D and 3D MLAD in that a systematically larger MLAD was found in the 2D
Fig. 3. A representative trial showing the deformation of the medial longitudinal arch (MLA) in degrees in relation to the vertical groundreaction force (GRF) in Newton. The circles indicate the different instants of the foot ground contact period where the MLA is calculated and the photos below illustrate the position of the foot at the specified time of the contact period.
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
403
Table 1 Descriptive data and reliability results. Parameter 3D 3D 2D 2D *
angle angle angle angle
1 4 1 4
(8) (8) (8) (8)
Test mean (SD)
Retest mean (SD)
Test–retest difference (SD)
ICC (2.1)
95% conf. Int.
SEM
154.9 163.1 150.3 159.8
154.9 163.0 150.5 160.0
0.0 0.1 0.2 0.2
0.95 0.95 0.95 0.95
0.92–0.97 0.92–0.97 0.91–0.97 0.92–0.97
3.4 3.5 3.4 3.8
(10.2) (11.5) (10.8)* (12.6)*
(10.5) (11.9) (11.4)* (13.0)*
(4.8) (5.0) (4.8) (5.4)
Denotes significant difference (p < 0.001) from respective 3D parameter.
Table 2 Descriptive data and reliability results of MLAD and ND. Parameter
Test mean (SD)
Retest mean (SD)
Test–retest difference (SD)
ICC (2.1)
95% conf. Int.
SEM
3D MLAD (8) 2D MLAD (8) ND (mm)
8.3 (2.8) 9.6 (3.4)* 7.2 (3.6)
8.1 (2.6) 9.5 (3.3)* 7.0 (3.9)
0.2 (1.1) 0.1 (1.9) 0.1 (2.4)
0.95 0.91 0.88
0.92–0.97 0.84–0.95 0.80–0.93
0.8 1.4 1.7
*
Denotes significant difference (p < 0.001) from 3D MLAD.
Table 3 Relationship between variables. Parameter
3D MLAD (8)
2D MLAD (8)
ND (mm)
3D MLAD (8)
–
2D MLAD (8)
–
0.76 p < 0.001 –
0.56 p < 0.001 0.61 p < 0.001
dynamic measurements (Table 2). The reliability measurements of static and dynamic deformation showed higher reliability in the dynamic tests than the static ND test (Table 2). The dynamic 3D MLAD showed less correlation with the static ND test (r = 0.56, p < 0.001) than with the 2D MLAD (r = 0.76, p < 0.001) (see Table 3). 5. Discussion In daily clinical examinations, time for laboratory testing is limited and demands for simple, reliable, and valid methods of testing are increasing. The present study investigated the reliability of sagittal plane variables of MLA during gait, using a simple modified 3D kinematic method. Three-dimensional studies have investigated the MLA either as the angle between three medially placed markers [2,26], as in the present study, or as angles between forefoot and hindfoot segments [27–29]. The range of MLA deformation in these studies ranges from about 58 [27] to about 128 [29]. The results on MLAD from this study (8.18 2.88) therefore concur with other studies, although the absolute values of MLA do not match the mentioned studies due to differences in marker placements. The reliability of 3D MLAD was excellent and comparable to the reliability measures reported in other 3D MLA studies [2,27]. The present method involved only three markers on each foot, and the total test time was less than 20 min. This method may therefore be more time-saving than other, more advanced, 3D analyses, which in some models require application of 17 or 19 markers [28,29]. The relative simplicity of the present model, compared to the advanced models, may increase its potential for clinical use. Besides reliability, the dynamic 3D MLAD test reflects how the supporting structures are responding to the external loads during dynamic activity. Since overloading injuries result from excess dynamic loading, dynamic tests like the 3D MLAD may be a better clinical predictor of overloading injury risk than static tests. The clinical significance of this MLAD method was evident in a study of MLAD differences between athletes with Medial Tibial Stress Syndrome and matched healthy athletes using the same procedures
as in the present study [15]. Even though other studies have shown good correlations between measures of MLA during gait midstance and static measures of MLA [22,23], the results of the present study show only a moderate correlation (Pearson’s, r = 0.56, p < 0.001) between the static ND test and the dynamic MLAD (Table 3). It could be speculated, that this lack of correlation may in part be explained by a higher loading of the MLA in the push-off situation compared to quiet standing. It has been shown, that the center of pressure is moving forward during the contact phase of walking, reaching a position under the forefoot during the push-off phase, and this increased external moment arm contributes greatly to the peak plantarflexor moment seen during the terminal stance phase in walking [30]. This is in contrast to quiet standing, where the center of pressure is located under the midfoot with much smaller plantarflexor moments. Assuming the foot as a two-segment structure in the sagittal plane with a hinge joint placed at the navicular bone between the forefoot and the hindfoot, the increased plantarflexor moment during push-off would correlate to an external moment attempting to dorsiflex the forefoot with respect to the hindfoot. The lack of good correlation between static and dynamic measures could thus be a result of this extra load being counteracted differently between participants. This would further imply that a static test measuring the ND might only have limited validity as predictor of MLAD during gait. Future investigations using kinetic foot models may assess the magnitude of these moments. In an attempt to make the method applicable in daily clinical practice, the 3D measurements were compared to the 2D measurements obtained by a single digital video camera placed perpendicularly to the direction of walking. There was a significant systematic difference between MLAD results from the two methods, which may be due to variations in foot progression angles observed in normal subjects. The repeatability of the 2D angles during gait was high and matched the 3D ICCs and SEMs. The correlation between the 3D measurement and the 2D measurement was higher than for the 3D measurement and the static ND test, but not excellent. This may be due to the slight variations of foot progression angles that are likely to occur during normal walking. However, the video based 2D measurements may be better than the static ND test in assessing the MLAD developed during gait, but may still have limited correlation to the probably more valid 3D MLAD, as the correlation coefficient (r = 0.76) suggests that 2D video measurements explain less than 60% of the variation of the 3D MLAD. Still, the 2D MLAD proves highly reproducible and may therefore be used in intervention studies, for example. Furthermore, this method may easily be applied in clinical settings without a biomechanical motion analysis system.
404
J. Bencke et al. / Gait & Posture 35 (2012) 400–404
The major limitation of the presented new method is that it only measures the longitudinal medial arch, although the MLAD may be a composite of deformations in all three planes. Therefore, e.g. excess pronation will not be measured isolated, and the clinical interpretation may therefore be performed with regard to this limitation. The dynamic methods examined furthermore demand access to a camera based movement analysis system with a forceplate or similar, which may be a practical limitation. In conclusion, the new and simple 3D method of measuring deformation of the MLA during gait showed excellent reliability, and was more reliable than a dynamic 2D method and a traditional static method, as was hypothezised. Furthermore, the results of the present study show that static measurements of MLAD such as the ND may have limited validity as predictors of MLAD during gait. Being relatively quick to perform in a biomechanical laboratory setting, the 3D method may be used in clinical examinations of unstable feet scheduled for surgery or as an outcome measure following surgery or rehabilitation. Further studies may evaluate the clinical validity of the presented method in terms of ability to detect clinically relevant intervention improvements or predicting risk of overuse injuries like medial tibial stress syndrome or plantar fasciitis. Acknowledgements The authors like to thank Derek Curtis, PT, MSc, for his valuable assistance in reading through the final version of the manuscript and correcting misspellings and grammatical errors. Conflict of interest There are no duplicate publication elsewhere of any part of the work, and there are no commercial interests in the results of the present study, or other conflicts of interest. References [1] Winson I, Lundberg A, Bylund C. The pattern of motion of the longitudinal arch of the foot. The Foot 1994;4:151–4. [2] Hunt AE, Smith RM, Torode M, Keenan AM. Inter-segment foot motion and ground reaction forces over the stance phase of walking. Clin Biomech (Bristol Avon) 2001;16(7):592–600. [3] Richie Jr DH. Biomechanics and clinical analysis of the adult acquired flatfoot. Clin Podiatr Med Surg 2007;24(4):617–44. vii.. [4] Thordarson DB, Schmotzer H, Chon J, Peters J. Dynamic support of the human longitudinal arch. A biomechanical evaluation. Clin Orthop Relat Res 1995;316:165–72. [5] Tao K, Ji WT, Wang DM, Wang CT, Wang X. Relative contributions of plantar fascia and ligaments on the arch static stability: a finite element study. Biomed Tech (Berl) 2010;55(5):265–71. [6] Fiolkowski P, Brunt D, Bishop M, Woo R, Horodyski M. Intrinsic pedal musculature support of the medial longitudinal arch: an electromyography study. J Foot Ankle Surg 2003;42(6):327–33.
[7] Saltzman CL, Nawoczenski DA, Talbot KD. Measurement of the medial longitudinal arch. Arch Phys Med Rehabil 1995;76(1):45–9. [8] Brody DM. Techniques in the evaluation and treatment of the injured runner. Orthop Clin North Am 1982;13(3):541–58. [9] Vinicombe A, Raspovic A, Menz HB. Reliability of navicular displacement measurement as a clinical indicator of foot posture. J Am Podiatr Med Assoc 2001;91(5):262–8. [10] Bennett JE, Reinking MF, Pluemer B, Pentel A, Seaton M, Killian C. Factors contributing to the development of medial tibial stress syndrome in high school runners. J Orthop Sports Phys Ther 2001;31(9):504–10. [11] Allen MK, Glasoe WM. Metrecom measurement of navicular drop in subjects with anterior cruciate ligament injury. J Athl Train 2000;35(4):403–6. [12] Nielsen RG, Rathleff MS, Moelgaard CM, Simonsen O, Kaalund S, Olesen CG, et al. Video based analysis of dynamic midfoot function and its relationship with Foot Posture Index scores. Gait Posture 2010;31(1):126–30. [13] Menz HB. Alternative techniques for the clinical assessment of foot pronation. J Am Podiatr Med Assoc 1998;88(3):119–29. [14] Couture CJ, Karlson KA. Tibial stress injuries: decisive diagnosis and treatment of ‘shin splints’. Phys Sportsmed 2002;30(6):29–36. [15] Bandholm T, Boysen L, Haugaard S, Zebis MK, Bencke J. Foot medial longitudinal-arch deformation during quiet standing and gait in subjects with medial tibial stress syndrome. J Foot Ankle Surg 2008;47(2):89–95. [16] Hubbard TJ, Carpenter EM, Cordova ML. Contributing factors to medial tibial stress syndrome: a prospective investigation. Med Sci Sports Exerc 2009;41 (3):490–6. [17] Myklebust G, Maehlum S, Holm I, Bahr R. A prospective cohort study of anterior cruciate ligament injuries in elite Norwegian team handball. Scand J Med Sci Sports 1998;8(3):149–53. [18] Jenkins WL, Killian CB, Williams III DS, Loudon J, Raedeke SG. Anterior cruciate ligament injury in female and male athletes: the relationship between foot structure and injury. J Am Podiatr Med Assoc 2007;97(5):371–6. [19] Devereaux MD, Lachmann SM. Patello-femoral arthralgia in athletes attending a sports injury clinic. Br J Sports Med 1984;18(1):18–21. [20] Hetsroni I, Finestone A, Milgrom C, Sira DB, Nyska M, Radeva-Petrova D, et al. A prospective biomechanical study of the association between foot pronation and the incidence of anterior knee pain among military recruits. J Bone Joint Surg Br 2006;88(7):905–8. [21] Duffey MJ, Martin DF, Cannon DW, Craven T, Messier SP. Etiologic factors associated with anterior knee pain in distance runners. Med Sci Sports Exerc 2000;32(11):1825–32. [22] McPoil TG, Cornwall MW. Use of the longitudinal arch angle to predict dynamic foot posture in walking. J Am Podiatr Med Assoc 2005;95(2): 114–20. [23] Franettovich MM, McPoil TG, Russell T, Skardoon G, Vicenzino B. The ability to predict dynamic foot posture from static measurements. J Am Podiatr Med Assoc 2007;97(2):115–20. [24] Cashmere T, Smith R, Hunt A. Medial longitudinal arch of the foot: stationary versus walking measures. Foot Ankle Int 1999;20(2):112–8. [25] Alkjaer T, Simonsen EB, Dyhre-Poulsen P. Comparison of inverse dynamics calculated by two- and three-dimensional models during walking. Gait Posture 2001;13(2):73–7. [26] Tome J, Nawoczenski DA, Flemister A, Houck J. Comparison of foot kinematics between subjects with posterior tibialis tendon dysfunction and healthy controls. J Orthop Sports Phys Ther 2006;36(9):635–44. [27] Nakamura H, Kakurai S. Relationship between the medial longitudinal arch movement and the pattern of rearfoot motion during the stance phase of walking. J Phys Ther Sci 2003;15(1):13–8. [28] MacWilliams BA, Cowley M, Nicholson DE. Foot kinematics and kinetics during adolescent gait. Gait Posture 2003;17(3):214–24. [29] Stebbins J, Harrington M, Thompson N, Zavatsky A, Theologis T. Repeatability of a model for measuring multi-segment foot kinematics in children. Gait Posture 2006;23(4):401–10. [30] Haim A, Rozen N, Wolf A. The influence of sagittal center of pressure offset on gait kinematics and kinetics. J Biomech 2010;43(5):969–77.