- Email: [email protected]
Comparison of footprint parameters calculated from static and dynamic footprints
The Foot (1999) 9, 145–149 © 1999 Harcourt Publishers Ltd
ORIGINAL ARTICLE
Comparison of footprint parameters calculated from static and dynamic footprints I. Mathieson, D. Upton, A. Birchenough Wales Centre for Podiatric Studies University of Wales Institute, Cardiff SUMMARY. In attempting to establish relationships between ‘foot-type’ and pathology, many researchers have chosen to use parameters calculated from static footprints to define study groups. In this study the ability of such static information to reflect the dynamic foot was investigated by comparing values for three parameters (Stahelis Arch Index, Chippaux-Smirak Index, Footprint Angle) calculated from static and dynamic prints. Three static and three dynamic footprints were obtained for 20 subjects. Reliability of the electronic data collection technique (P<0.001) was good for all three parameters, while linear parameters displayed greater between-print reliability (P<0.001) than the angular. Parameters calculated from dynamic prints were found to differ significantly (P<0.05) from those calculated from static prints, although good correlation’s between the two states (P<0.001) indicate a general dynamic increase of 28%. Further, the Stahelis arch and Chippaux-Smirak indices, both claiming to measure the same condition, correlated very well with each other (P<0.001). Continued investigation of footprint parameters appears warranted. © 1999 Harcourt Publishers Ltd
conducted using such parameters, and also limit the potential for meta-analysis.6 The suggestion that foottype parameters offer advantages over lengthy clinical examinations for epidemiological studies requiring large numbers of subjects7 fails to justify adequately their continued use. Foot-type parameters typically used include the Arch Index,4 Stahelis Arch Index8 and the ChippauxSmirak Index,9 all of which are calculated from static footprints. These parameters operate on the premise that the footprint responds predictably to variations in the structure of the medial longitudinal arch, which has been described as one of the most important structural characteristics of the foot.4,10 Qamra et al.11 conducted an evaluation of footprints, stating that forefoot and heel areas consistently show maximum ground contact with progressive changes in arch height, restricting changes in the footprint to the central portion. Several measures appear consistent with this theory, as they involve a direct comparison of the contact areas in either the forefoot or heel with the midfoot. For example, the arch index of Cavanagh and Rodgers4 involves comparison of the contact area of the midfoot with the contact area of the rest of the print minus the toes. The same rationale has resulted in more convenient indices which are less time-consuming to calculate. These involve comparison of the midfoot width, firstly with the forefoot width, to calculate the Chippaux-Smirak index9 and secondly with the rearfoot width, to calculate Stahelis Arch Index.8 The Footprint
INTRODUCTION Foot-type is a general term which has been used extensively to describe a range of architectural features of the foot thought to provide clues regarding dynamic function. Dynamic foot function in turn is perceived to be related to a variety of musculoskeletal symptoms, including shin splints,1 running injuries2 and plantar heel pain.3 Commonly, such studies have progressed by comparing the presence, absence or magnitude of foot-type parameters between groups with and without the pathology under investigation. Despite the widespread use of these parameters, and the acknowledgement that certain foot-types are recognized as separate clinical entities (i.e. Pes Planus and Pes Cavus),4 it has been conceded that an objective, quantitative measure of foot-type remains elusive. This absence of an absolute measure of foottype has led to considerable variation in the choice of measure used as the exposure variable for studies concerning foot-type. This is in contrast to the suggestion that if true relationships between foottype and pathology are to be identified, it is first necessary to have a valid system of classification to allow accurate recognition of each state.5 This issue appears to compromise the validity of research
Correspondence to I. Mathieson, Wales Centre for Podiatric Studies, University of Wales Institute, Cardiff, Western Avenue, Cardiff, Wales CF5 2YB, UK. Tel: +44 (0)1222 506 864; Fax: +44 (0)1222 506 980; E-mail:[email protected] 145
146
The Foot from static prints for the purpose of standardization would have been compromised if static prints were found to differ significantly from dynamic. The ability of static parameters to predict the dynamic would appear to be a fundamental, although not exclusive, pre-requisite for validity. This study set out to investigate the agreement between the Staheli Arch Index, the ChippauxSmirak index and the Footprint Angle calculated from static vs dynamic prints.
METHOD A convenience sample of 20 university student and staff volunteers was used: all participants satisfied the following inclusion criteria: • No history of, or visually apparent, gait disturbance • No history of trauma, including repetitive minor trauma such as ankle sprains history of any systemic illness which may • No influence gait.
Fig. 1 Calculation methods for Stahelis arch index (SAI), Chippaux-Smirak index (CSI) and Footprint angle (FPA). SAI involves measuring the width of the foot in the area of the arch (line B) and the rearfoot (line C), and dividing the former by the latter. This is similar to the CSI, where the width of the forefoot (line A) and the heel (line C) are measured, the latter divided by the former and expressed as a percentage. The FPA (angle α) is measured as the angle between the medial reference line and a line connecting the most medial and anterior aspect of the medial longitudinal arch.
Angle12,13 may also demonstrate potential as it appears to examine directly the response of the midfoot, through examination of the angle between a medial reference line (a line connecting the most medial margins of the forefoot and rearfoot areas) and the most anterior and lateral margin of the medial longitudinal arch, which may also change orientation with changes in arch height. These techniques are illustrated in Figure 1. It has been suggested that measures of foot-type be evaluated with special regard to the dynamic rather than the static situation.14,15 This suggestion appears important since it is dynamic foot function that is believed to be related to pathology. However, currently such parameters are calculated routinely from static prints, which may not represent the true condition of the foot under dynamic conditions, where variables such as angle and base of gait may act through established compensatory mechanisms to influence internal and external architecture.16 Cavanagh and Rodgers4 previously investigated the relationship between Arch Indices calculated from static, walking and running conditions, and despite noting large variation, found an average increase of 10% between static and running conditions. The recommendation that the index should be calculated The Foot (1999) 9, 145–149
The sample comprised 14 females and 6 males, average age 25 (±9) and average body mass index of 23 (±3.2). Static and dynamic footprints, three of each, were obtained for each subject using the Musgrave Footprint System (Musgrave Systems Ltd, Wrexham, UK), which allowed an image of the plantar surface of the foot to be easily captured. The Musgrave plates were embedded mid-way along an 8 m walkway. Electronic data capture was chosen due to the ease of subsequent analysis using standard Musgrave software (Footstats). Static print collection method The subject was asked to stand barefoot one step behind the first plate on the walkway with his right foot lined up with the plate. When ready they took one step forward, placing their right foot centrally within the plate. The left foot was then brought alongside the right, and the subject was allowed to vary the position of this foot until a comfortable stance position was attained. With each subject standing relaxed, data capture was initiated. The Musgrave Footprint was set to record data for one second. This procedure was repeated until three satisfactory prints were obtained for each subject. Dynamic print collection method Subjects were positioned at the start of the walkway, and were instructed to walk up and down five times, ignoring the plates, in an attempt to naturalize their gait patterns. The subjects then returned to the start of the walkway, lined up their right foot with the right plate, and were instructed to begin each walk with a right step. Another walk was undertaken, still ignoring the plates, and an observer advised of any changes © 1999 Harcourt Publishers Ltd
Comparison of footprint parameters Table 1
147
Reliability of electronic footprint analysis technique (Pearson’s rho)
First evaluation vs Second evaluation
Footprint Angle
Chippaux-Smirak
Stahelis Arch
0.94
0.93
0.96
Correlation’s between two separate evaluations of the same footprint. Values consistently above 0.9 equates with significance at the P<0.001 level.
Table 2 Reliability of static and dynamic footprinting techniques (Pearson’s rho) Static footprints Analysis 1 vs 2 1 vs 3 2 vs 3
Dynamic footprints
Footprint Angle
ChippauxSmirak
Stahelis Arch
Footprint Angle
ChippauxSmirak
Stahelis Arch
0.51* 0.29 0.33
0.92* 0.95* 0.92*
0.91* 0.94* 0.91*
0.78* 0.63* 0.66*
0.8* 0.8* 0.9*
0.86* 0.81* 0.91*
All correlation’s marked * are significant at the P<0.001 level. With the calculation method established as reliable, any change of value between separate prints of the same foot would suggest that the dimensions of the foot differ significantly between footprints. However, comparison of values derived from three separate prints indicates that linear based parameters do not differ significantly between footprints for static and dynamic conditions. The footprint angle was found to differ between separate static prints.
to the starting position to facilitate a central strike. This process was repeated until appropriate starting positions were identified. The subjects were instructed to begin walking up and down, being unaware of which footprints were being recorded. They were instructed to stop only when three satisfactory prints had been obtained. Any prints which failed to capture a clear image of the entire plantar aspect of the foot were rejected. The study yielded three static and three dynamic prints of the right foot for each subject, resulting in 120 prints. A single tester evaluated all prints, calculating Footprint Angle, and maximal forefoot width, minimal midfoot width and maximal rearfoot width to allow the calculation of the Chippaux-Smirak index and Stahelis Arch Index. The data was found to approximate to a normal distribution, therefore, parametric analysis techniques were employed.
RESULTS Reliability of electronic footprint analysis technique Firstly, the reliability of the measurement technique used to extract the raw data used to calculate the chosen parameters was investigated. This was achieved by calculating the parameters from the same prints on two separate occasions and examining agreement between them. The first static print for each subject was chosen for this purpose, and the two occasions were 4 days apart. Comparison of parameters from S1a (first evaluation of static print 1) and S1b (second evaluation of static print 1) was performed using Pearson’s correlation coefficients, and are shown in Table 1. In all cases, agreement between first and second assessments of the same prints was excellent, being significant at the P<0.001 level. © 1999 Harcourt Publishers Ltd
Reliability of static and dynamic footprinting techniques The reliability of the footprinting techniques was evaluated by examining agreement between parameters calculated from different prints of the same foot. Separate analysis of static and dynamic reliability was performed using Pearson’s correlation coefficients. Results are shown in Table 2. The footprint angle displayed unacceptable reliability when calculated from static prints, and therefore, it was excluded from further analysis. Comparison between statically and dynamically calculated parameters The relationship between parameters calculated from static and dynamic prints was examined by comparing the mean static and mean dynamic values of each parameter. Firstly, paired t-tests were performed to investigate differences between the two states, followed by Pearson’s correlation coefficients to investigate the consistency of any differences.
Chippaux-Smirak Index calculated from static and dynamic footprints Analysis demonstrated a difference between the two states (t=2.5, df=19, P=0.02) with dynamic prints (mean=24.7 S.D.=13.2) being higher than the static (mean=21.03 S.D.=14.34), and a mean difference of 3.7 (95% CI = –6.7: –0.65). A Pearson’s correlation coefficient of 0.89 (P<0.001) indicates that these differences are consistent. Expression of the difference as a percentage of the static value revealed a mean increase of 25.56±43.36, indicating large variation in the magnitude of this increase between static and The Foot (1999) 9, 145–149
148
The Foot
Fig. 2 Correlation between mean static and mean dynamic Chippaux-Smirak Index.
Fig. 3 Correlation between mean static and mean dynamic Stahelis Arch Index.
dynamic prints. The correlation between static and dynamic is illustrated in Figure 2.
essential, although not exclusive, assumption central to parameter validity regards the ability of the static print to reflect the dynamic. This study indicates that statically calculated footprint parameters do infer something of the dynamic situation. However, although a general increase of approximately 25% (Chippaux-Smirak) and 28% (Stahelis Arch Index) was found between static and dynamic states, large individual variations prevent any inferences from being made. Statically calculated parameters must be viewed with caution, as they appear to inconsistently predict even the dynamic dimensions of the foot. The relationship to dynamic motion – the variable perceived to be related to the development of pathology – remains unclear. Acceptance of results pertaining to correlation’s between static and dynamic prints is dependant upon establishing the reliability of the data collection techniques. As such, the reliability of both static and dynamic print collection techniques and the software used to examine the footprints required evaluation. Firstly, the reliability of the electronic footprint assessment technique, involving the use of Musgrave Footprint software, was examined through correlation between two separate analyses of the first set of static prints. Correlation’s consistently above 0.9 (P<0.001) suggest that this software is reliable, justifying its use, and allowing it to be recommended. Secondly, the reliability of the static and dynamic print collection techniques was examined by correlating the same parameter calculated from three separate prints obtained in the same state. Although correlation’s were excellent for the Chippaux-Smirak and Stahelis Arch indices for both static and dynamic states, the footprint angle failed to show such convincing reliability. Correlation’s were generally lower, and were non-significant for two of three tests performed between static prints. It is interesting that dynamic prints showed greater reliability than the static, given the inherent variability of the dynamic situation, which could have been expected to produce lower reliability for the dynamic. Although good reliability coefficients have been reported previously14 these figures relate to separate analyses of the same prints, both by single and multiple investigators. The likelihood that different prints of the same foot would
Stahelis Arch Index calculated from static and dynamic footprints Analysis again demonstrated a difference between the two states (t=2.6, df=19, P=0.018) with the dynamic (mean=0.38 S.D.=0.21) greater than the static (mean=0.32 S.D.=0.21) and a mean difference of 5.85E-02 (95% CI = –0.1: –1.14E-02). A Pearson’s correlation of 0.88 (P<0.001) again demonstrated that these differences appear to be consistent. Expression of the mean difference as a percentage of the static value revealed a mean increase of 27.95±57.65, indicating large variation in the magnitude of this increase between static and dynamic prints. Again, the correlation between static and dynamic is illustrated in Figure 3. Comparison of separate parameters A final correlation was performed to examine the agreement between the two separate parameters which claim to be measuring the same condition. This was achieved by correlating the mean static Chippaux-Smirak and mean static Stahelis Arch indices, and the mean dynamic Chippaux-Smirak and mean dynamic Stahelis Arch Indices. Correlation’s of 0.99 (P<0.001) for the static and 0.98 (P<0.001) for the dynamic indicate that the parameters do appear to be sensitive to the same feature.
DISCUSSION The concept that parameters calculated from static footprints can accurately be used to categorize feet according to structural characteristics has led to the widespread use of such indices for research purposes. This use has persisted despite an absence of supportive evidence, and acknowledgement of the fact that examination of footprints offers only a possibility of an indirect statement about the medial longitudinal arch4. Since the structural characteristics are directly perceived to influence dynamic behaviour, one The Foot (1999) 9, 145–149
© 1999 Harcourt Publishers Ltd
Comparison of footprint parameters return the same value was not investigated. Poor between-print reliability of the Footprint Angle is felt to be an important finding. Simply dismissing this parameter may not be warranted, however, since poor reliability may be a consequence of high sensitivity to the discrete changes in arch configuration which could be attributed to postural sway. Inclusion of the Footprint Angle in a validity study may be warranted. Additional methodological variables potentially influencing results include the use of forceplates embedded in a walkway, and the associated problem of ‘targeting’, which is often felt to reduce the validity of the resulting dynamic prints. This issue has been investigated at length,17–20 with the ‘conventional wisdom’ that targeting has a deleterious effect remaining unsubstantiated.17 Contrary to the belief that targeting has a negative effect, Harrison and Folland19 investigated the effects of five different protocols for the collection of dynamic prints by examining the spatial distribution of forces on seven discrete areas of the foot and found minimal differences. Such evidence suggests that forceplates yield valid information, and supports their use. CONCLUSION The ability of parameters calculated from static footprints to relate to the dynamic foot appears limited. Although parameters calculated from dynamic prints are generally higher, the magnitude of this increase varies considerably. The continued use of footprint parameters needs to be substantiated by a validity study. ACKNOWLEDGEMENTS The authors wish to thank Mr Declan O’Doherty and Mr Richard Callaghan for their valuable contributions, and N. Mathieson for preparing the illustrations. REFERENCES 1. Wen D Y, Puffer J C, Schmalzried T P, Lower Extremity alignment and risk of overuse injuries in runners. Med Sci Sp Exerc 1997; 29: 1291–1298.
© 1999 Harcourt Publishers Ltd
149
2. Busseuil C, Freychat P, Guedj E B, Lacour J R Forefootrearfoot orientation and traumatic risk for runners. Foot & Ankle Int 1998; 19: 32–37. 3. Prichasuk S, Subhadrabandhu T. The relationship of Pes Planus and calcaneal spur to plantar heel pain. Clin Orthop Rel Res 1994; 306: 192–196. 4. Cavanagh P R, Rodgers M M, The arch index: a useful measure from footprints. J Biomechanics 1987; 20: 547–551. 5. Rose G K, Pes Planus. In: Jahss M (Ed). Disorders of the Foot, 2nd ed. Philadelphia: Saunders, 1991; 486–520. 6. Ilahi O A, Kohl H W. Lower extremity morphology and alignment and risk of overuse injury. Clin J Sport Med 1998; 8: 38–42. 7. Cowan D N, Robinson J R, Jones B H, Polly D W, Berry B H Consistency of visual assessments of arch height among clinicians. Foot & Ankle International 1994; 15: 213–217 8. Staheli L T, Chew D E, Corbett M The longitudinal arch: A survey of 882 feet in normal children and adults. J Bone Joint Surg 1987; 69-A 426–428. 9. Forriol F, Pascual J Footprint analysis between the ages of 3 and 17 years of age. Foot & Ankle 1990; 11: 101–104. 10. McCrory J L, Young M J, Boulton A J M, Cavanagh P R Arch index as a predictor of arch height. Foot 1997; 7: 79–81. 11. Qamra S R, Deodhar S D, Jit I Podographical and metrical study for Pes Planus in a Northwestern Indian population. Human Biology 1980; 52: 435–445. 12. Schwartz L, Britten R H, ThompsonL R Studies in Physical Development and Posture. US Public Health Bulletin 1928; 179: 23. 13. Clarke H H An Objective Method of Measuring the Height of the Longitudinal Arch in Foot Examinations. Res Quarterly 1933; 4: 99–107. 14. Cureton T K The validity of footprints as a measure of the vertical height of the arch and functional efficiency of the foot. Res Quarterly 1935; 6: 70–80. 15. Rose G K, Welton E A, Marshall T The diagnosis of flatfoot in the child. J Bone Joint Surg 1985; 67-B pp 71–78. 16. Michaud T C Foot orthoses and other forms of conservative foot care. Massachusetts 1998. 17. Sanderson D J, Franks I M, Elliott D The effects of targeting on the ground reaction forces during level walking. Human Movement Science 1993; 12: 327–337. 18. Grabiner M D, Feuerbach J W, Lundin T M, Davis B L Visual guidance to force plates does not influence ground reaction force variability. J Biomechanics 1995; 28: 1115–1117. 19. Harrison A J, Folland J P Investigation of gait protocols for plantar pressure measurement of non-pathological subjects using a dynamic pedobarograph. Gait & Posture 1997; 6: 50–55. 20. Rietdyk S, Patla A E Does the step length requirement in the subsequent step influence the strategies used for step length regulation in the current step? Human Movement Science 1994; 13: 109–127.
The Foot (1999) 9, 145–149
ORIGINAL ARTICLE
Comparison of footprint parameters calculated from static and dynamic footprints I. Mathieson, D. Upton, A. Birchenough Wales Centre for Podiatric Studies University of Wales Institute, Cardiff SUMMARY. In attempting to establish relationships between ‘foot-type’ and pathology, many researchers have chosen to use parameters calculated from static footprints to define study groups. In this study the ability of such static information to reflect the dynamic foot was investigated by comparing values for three parameters (Stahelis Arch Index, Chippaux-Smirak Index, Footprint Angle) calculated from static and dynamic prints. Three static and three dynamic footprints were obtained for 20 subjects. Reliability of the electronic data collection technique (P<0.001) was good for all three parameters, while linear parameters displayed greater between-print reliability (P<0.001) than the angular. Parameters calculated from dynamic prints were found to differ significantly (P<0.05) from those calculated from static prints, although good correlation’s between the two states (P<0.001) indicate a general dynamic increase of 28%. Further, the Stahelis arch and Chippaux-Smirak indices, both claiming to measure the same condition, correlated very well with each other (P<0.001). Continued investigation of footprint parameters appears warranted. © 1999 Harcourt Publishers Ltd
conducted using such parameters, and also limit the potential for meta-analysis.6 The suggestion that foottype parameters offer advantages over lengthy clinical examinations for epidemiological studies requiring large numbers of subjects7 fails to justify adequately their continued use. Foot-type parameters typically used include the Arch Index,4 Stahelis Arch Index8 and the ChippauxSmirak Index,9 all of which are calculated from static footprints. These parameters operate on the premise that the footprint responds predictably to variations in the structure of the medial longitudinal arch, which has been described as one of the most important structural characteristics of the foot.4,10 Qamra et al.11 conducted an evaluation of footprints, stating that forefoot and heel areas consistently show maximum ground contact with progressive changes in arch height, restricting changes in the footprint to the central portion. Several measures appear consistent with this theory, as they involve a direct comparison of the contact areas in either the forefoot or heel with the midfoot. For example, the arch index of Cavanagh and Rodgers4 involves comparison of the contact area of the midfoot with the contact area of the rest of the print minus the toes. The same rationale has resulted in more convenient indices which are less time-consuming to calculate. These involve comparison of the midfoot width, firstly with the forefoot width, to calculate the Chippaux-Smirak index9 and secondly with the rearfoot width, to calculate Stahelis Arch Index.8 The Footprint
INTRODUCTION Foot-type is a general term which has been used extensively to describe a range of architectural features of the foot thought to provide clues regarding dynamic function. Dynamic foot function in turn is perceived to be related to a variety of musculoskeletal symptoms, including shin splints,1 running injuries2 and plantar heel pain.3 Commonly, such studies have progressed by comparing the presence, absence or magnitude of foot-type parameters between groups with and without the pathology under investigation. Despite the widespread use of these parameters, and the acknowledgement that certain foot-types are recognized as separate clinical entities (i.e. Pes Planus and Pes Cavus),4 it has been conceded that an objective, quantitative measure of foot-type remains elusive. This absence of an absolute measure of foottype has led to considerable variation in the choice of measure used as the exposure variable for studies concerning foot-type. This is in contrast to the suggestion that if true relationships between foottype and pathology are to be identified, it is first necessary to have a valid system of classification to allow accurate recognition of each state.5 This issue appears to compromise the validity of research
Correspondence to I. Mathieson, Wales Centre for Podiatric Studies, University of Wales Institute, Cardiff, Western Avenue, Cardiff, Wales CF5 2YB, UK. Tel: +44 (0)1222 506 864; Fax: +44 (0)1222 506 980; E-mail:[email protected] 145
146
The Foot from static prints for the purpose of standardization would have been compromised if static prints were found to differ significantly from dynamic. The ability of static parameters to predict the dynamic would appear to be a fundamental, although not exclusive, pre-requisite for validity. This study set out to investigate the agreement between the Staheli Arch Index, the ChippauxSmirak index and the Footprint Angle calculated from static vs dynamic prints.
METHOD A convenience sample of 20 university student and staff volunteers was used: all participants satisfied the following inclusion criteria: • No history of, or visually apparent, gait disturbance • No history of trauma, including repetitive minor trauma such as ankle sprains history of any systemic illness which may • No influence gait.
Fig. 1 Calculation methods for Stahelis arch index (SAI), Chippaux-Smirak index (CSI) and Footprint angle (FPA). SAI involves measuring the width of the foot in the area of the arch (line B) and the rearfoot (line C), and dividing the former by the latter. This is similar to the CSI, where the width of the forefoot (line A) and the heel (line C) are measured, the latter divided by the former and expressed as a percentage. The FPA (angle α) is measured as the angle between the medial reference line and a line connecting the most medial and anterior aspect of the medial longitudinal arch.
Angle12,13 may also demonstrate potential as it appears to examine directly the response of the midfoot, through examination of the angle between a medial reference line (a line connecting the most medial margins of the forefoot and rearfoot areas) and the most anterior and lateral margin of the medial longitudinal arch, which may also change orientation with changes in arch height. These techniques are illustrated in Figure 1. It has been suggested that measures of foot-type be evaluated with special regard to the dynamic rather than the static situation.14,15 This suggestion appears important since it is dynamic foot function that is believed to be related to pathology. However, currently such parameters are calculated routinely from static prints, which may not represent the true condition of the foot under dynamic conditions, where variables such as angle and base of gait may act through established compensatory mechanisms to influence internal and external architecture.16 Cavanagh and Rodgers4 previously investigated the relationship between Arch Indices calculated from static, walking and running conditions, and despite noting large variation, found an average increase of 10% between static and running conditions. The recommendation that the index should be calculated The Foot (1999) 9, 145–149
The sample comprised 14 females and 6 males, average age 25 (±9) and average body mass index of 23 (±3.2). Static and dynamic footprints, three of each, were obtained for each subject using the Musgrave Footprint System (Musgrave Systems Ltd, Wrexham, UK), which allowed an image of the plantar surface of the foot to be easily captured. The Musgrave plates were embedded mid-way along an 8 m walkway. Electronic data capture was chosen due to the ease of subsequent analysis using standard Musgrave software (Footstats). Static print collection method The subject was asked to stand barefoot one step behind the first plate on the walkway with his right foot lined up with the plate. When ready they took one step forward, placing their right foot centrally within the plate. The left foot was then brought alongside the right, and the subject was allowed to vary the position of this foot until a comfortable stance position was attained. With each subject standing relaxed, data capture was initiated. The Musgrave Footprint was set to record data for one second. This procedure was repeated until three satisfactory prints were obtained for each subject. Dynamic print collection method Subjects were positioned at the start of the walkway, and were instructed to walk up and down five times, ignoring the plates, in an attempt to naturalize their gait patterns. The subjects then returned to the start of the walkway, lined up their right foot with the right plate, and were instructed to begin each walk with a right step. Another walk was undertaken, still ignoring the plates, and an observer advised of any changes © 1999 Harcourt Publishers Ltd
Comparison of footprint parameters Table 1
147
Reliability of electronic footprint analysis technique (Pearson’s rho)
First evaluation vs Second evaluation
Footprint Angle
Chippaux-Smirak
Stahelis Arch
0.94
0.93
0.96
Correlation’s between two separate evaluations of the same footprint. Values consistently above 0.9 equates with significance at the P<0.001 level.
Table 2 Reliability of static and dynamic footprinting techniques (Pearson’s rho) Static footprints Analysis 1 vs 2 1 vs 3 2 vs 3
Dynamic footprints
Footprint Angle
ChippauxSmirak
Stahelis Arch
Footprint Angle
ChippauxSmirak
Stahelis Arch
0.51* 0.29 0.33
0.92* 0.95* 0.92*
0.91* 0.94* 0.91*
0.78* 0.63* 0.66*
0.8* 0.8* 0.9*
0.86* 0.81* 0.91*
All correlation’s marked * are significant at the P<0.001 level. With the calculation method established as reliable, any change of value between separate prints of the same foot would suggest that the dimensions of the foot differ significantly between footprints. However, comparison of values derived from three separate prints indicates that linear based parameters do not differ significantly between footprints for static and dynamic conditions. The footprint angle was found to differ between separate static prints.
to the starting position to facilitate a central strike. This process was repeated until appropriate starting positions were identified. The subjects were instructed to begin walking up and down, being unaware of which footprints were being recorded. They were instructed to stop only when three satisfactory prints had been obtained. Any prints which failed to capture a clear image of the entire plantar aspect of the foot were rejected. The study yielded three static and three dynamic prints of the right foot for each subject, resulting in 120 prints. A single tester evaluated all prints, calculating Footprint Angle, and maximal forefoot width, minimal midfoot width and maximal rearfoot width to allow the calculation of the Chippaux-Smirak index and Stahelis Arch Index. The data was found to approximate to a normal distribution, therefore, parametric analysis techniques were employed.
RESULTS Reliability of electronic footprint analysis technique Firstly, the reliability of the measurement technique used to extract the raw data used to calculate the chosen parameters was investigated. This was achieved by calculating the parameters from the same prints on two separate occasions and examining agreement between them. The first static print for each subject was chosen for this purpose, and the two occasions were 4 days apart. Comparison of parameters from S1a (first evaluation of static print 1) and S1b (second evaluation of static print 1) was performed using Pearson’s correlation coefficients, and are shown in Table 1. In all cases, agreement between first and second assessments of the same prints was excellent, being significant at the P<0.001 level. © 1999 Harcourt Publishers Ltd
Reliability of static and dynamic footprinting techniques The reliability of the footprinting techniques was evaluated by examining agreement between parameters calculated from different prints of the same foot. Separate analysis of static and dynamic reliability was performed using Pearson’s correlation coefficients. Results are shown in Table 2. The footprint angle displayed unacceptable reliability when calculated from static prints, and therefore, it was excluded from further analysis. Comparison between statically and dynamically calculated parameters The relationship between parameters calculated from static and dynamic prints was examined by comparing the mean static and mean dynamic values of each parameter. Firstly, paired t-tests were performed to investigate differences between the two states, followed by Pearson’s correlation coefficients to investigate the consistency of any differences.
Chippaux-Smirak Index calculated from static and dynamic footprints Analysis demonstrated a difference between the two states (t=2.5, df=19, P=0.02) with dynamic prints (mean=24.7 S.D.=13.2) being higher than the static (mean=21.03 S.D.=14.34), and a mean difference of 3.7 (95% CI = –6.7: –0.65). A Pearson’s correlation coefficient of 0.89 (P<0.001) indicates that these differences are consistent. Expression of the difference as a percentage of the static value revealed a mean increase of 25.56±43.36, indicating large variation in the magnitude of this increase between static and The Foot (1999) 9, 145–149
148
The Foot
Fig. 2 Correlation between mean static and mean dynamic Chippaux-Smirak Index.
Fig. 3 Correlation between mean static and mean dynamic Stahelis Arch Index.
dynamic prints. The correlation between static and dynamic is illustrated in Figure 2.
essential, although not exclusive, assumption central to parameter validity regards the ability of the static print to reflect the dynamic. This study indicates that statically calculated footprint parameters do infer something of the dynamic situation. However, although a general increase of approximately 25% (Chippaux-Smirak) and 28% (Stahelis Arch Index) was found between static and dynamic states, large individual variations prevent any inferences from being made. Statically calculated parameters must be viewed with caution, as they appear to inconsistently predict even the dynamic dimensions of the foot. The relationship to dynamic motion – the variable perceived to be related to the development of pathology – remains unclear. Acceptance of results pertaining to correlation’s between static and dynamic prints is dependant upon establishing the reliability of the data collection techniques. As such, the reliability of both static and dynamic print collection techniques and the software used to examine the footprints required evaluation. Firstly, the reliability of the electronic footprint assessment technique, involving the use of Musgrave Footprint software, was examined through correlation between two separate analyses of the first set of static prints. Correlation’s consistently above 0.9 (P<0.001) suggest that this software is reliable, justifying its use, and allowing it to be recommended. Secondly, the reliability of the static and dynamic print collection techniques was examined by correlating the same parameter calculated from three separate prints obtained in the same state. Although correlation’s were excellent for the Chippaux-Smirak and Stahelis Arch indices for both static and dynamic states, the footprint angle failed to show such convincing reliability. Correlation’s were generally lower, and were non-significant for two of three tests performed between static prints. It is interesting that dynamic prints showed greater reliability than the static, given the inherent variability of the dynamic situation, which could have been expected to produce lower reliability for the dynamic. Although good reliability coefficients have been reported previously14 these figures relate to separate analyses of the same prints, both by single and multiple investigators. The likelihood that different prints of the same foot would
Stahelis Arch Index calculated from static and dynamic footprints Analysis again demonstrated a difference between the two states (t=2.6, df=19, P=0.018) with the dynamic (mean=0.38 S.D.=0.21) greater than the static (mean=0.32 S.D.=0.21) and a mean difference of 5.85E-02 (95% CI = –0.1: –1.14E-02). A Pearson’s correlation of 0.88 (P<0.001) again demonstrated that these differences appear to be consistent. Expression of the mean difference as a percentage of the static value revealed a mean increase of 27.95±57.65, indicating large variation in the magnitude of this increase between static and dynamic prints. Again, the correlation between static and dynamic is illustrated in Figure 3. Comparison of separate parameters A final correlation was performed to examine the agreement between the two separate parameters which claim to be measuring the same condition. This was achieved by correlating the mean static Chippaux-Smirak and mean static Stahelis Arch indices, and the mean dynamic Chippaux-Smirak and mean dynamic Stahelis Arch Indices. Correlation’s of 0.99 (P<0.001) for the static and 0.98 (P<0.001) for the dynamic indicate that the parameters do appear to be sensitive to the same feature.
DISCUSSION The concept that parameters calculated from static footprints can accurately be used to categorize feet according to structural characteristics has led to the widespread use of such indices for research purposes. This use has persisted despite an absence of supportive evidence, and acknowledgement of the fact that examination of footprints offers only a possibility of an indirect statement about the medial longitudinal arch4. Since the structural characteristics are directly perceived to influence dynamic behaviour, one The Foot (1999) 9, 145–149
© 1999 Harcourt Publishers Ltd
Comparison of footprint parameters return the same value was not investigated. Poor between-print reliability of the Footprint Angle is felt to be an important finding. Simply dismissing this parameter may not be warranted, however, since poor reliability may be a consequence of high sensitivity to the discrete changes in arch configuration which could be attributed to postural sway. Inclusion of the Footprint Angle in a validity study may be warranted. Additional methodological variables potentially influencing results include the use of forceplates embedded in a walkway, and the associated problem of ‘targeting’, which is often felt to reduce the validity of the resulting dynamic prints. This issue has been investigated at length,17–20 with the ‘conventional wisdom’ that targeting has a deleterious effect remaining unsubstantiated.17 Contrary to the belief that targeting has a negative effect, Harrison and Folland19 investigated the effects of five different protocols for the collection of dynamic prints by examining the spatial distribution of forces on seven discrete areas of the foot and found minimal differences. Such evidence suggests that forceplates yield valid information, and supports their use. CONCLUSION The ability of parameters calculated from static footprints to relate to the dynamic foot appears limited. Although parameters calculated from dynamic prints are generally higher, the magnitude of this increase varies considerably. The continued use of footprint parameters needs to be substantiated by a validity study. ACKNOWLEDGEMENTS The authors wish to thank Mr Declan O’Doherty and Mr Richard Callaghan for their valuable contributions, and N. Mathieson for preparing the illustrations. REFERENCES 1. Wen D Y, Puffer J C, Schmalzried T P, Lower Extremity alignment and risk of overuse injuries in runners. Med Sci Sp Exerc 1997; 29: 1291–1298.
© 1999 Harcourt Publishers Ltd
149
2. Busseuil C, Freychat P, Guedj E B, Lacour J R Forefootrearfoot orientation and traumatic risk for runners. Foot & Ankle Int 1998; 19: 32–37. 3. Prichasuk S, Subhadrabandhu T. The relationship of Pes Planus and calcaneal spur to plantar heel pain. Clin Orthop Rel Res 1994; 306: 192–196. 4. Cavanagh P R, Rodgers M M, The arch index: a useful measure from footprints. J Biomechanics 1987; 20: 547–551. 5. Rose G K, Pes Planus. In: Jahss M (Ed). Disorders of the Foot, 2nd ed. Philadelphia: Saunders, 1991; 486–520. 6. Ilahi O A, Kohl H W. Lower extremity morphology and alignment and risk of overuse injury. Clin J Sport Med 1998; 8: 38–42. 7. Cowan D N, Robinson J R, Jones B H, Polly D W, Berry B H Consistency of visual assessments of arch height among clinicians. Foot & Ankle International 1994; 15: 213–217 8. Staheli L T, Chew D E, Corbett M The longitudinal arch: A survey of 882 feet in normal children and adults. J Bone Joint Surg 1987; 69-A 426–428. 9. Forriol F, Pascual J Footprint analysis between the ages of 3 and 17 years of age. Foot & Ankle 1990; 11: 101–104. 10. McCrory J L, Young M J, Boulton A J M, Cavanagh P R Arch index as a predictor of arch height. Foot 1997; 7: 79–81. 11. Qamra S R, Deodhar S D, Jit I Podographical and metrical study for Pes Planus in a Northwestern Indian population. Human Biology 1980; 52: 435–445. 12. Schwartz L, Britten R H, ThompsonL R Studies in Physical Development and Posture. US Public Health Bulletin 1928; 179: 23. 13. Clarke H H An Objective Method of Measuring the Height of the Longitudinal Arch in Foot Examinations. Res Quarterly 1933; 4: 99–107. 14. Cureton T K The validity of footprints as a measure of the vertical height of the arch and functional efficiency of the foot. Res Quarterly 1935; 6: 70–80. 15. Rose G K, Welton E A, Marshall T The diagnosis of flatfoot in the child. J Bone Joint Surg 1985; 67-B pp 71–78. 16. Michaud T C Foot orthoses and other forms of conservative foot care. Massachusetts 1998. 17. Sanderson D J, Franks I M, Elliott D The effects of targeting on the ground reaction forces during level walking. Human Movement Science 1993; 12: 327–337. 18. Grabiner M D, Feuerbach J W, Lundin T M, Davis B L Visual guidance to force plates does not influence ground reaction force variability. J Biomechanics 1995; 28: 1115–1117. 19. Harrison A J, Folland J P Investigation of gait protocols for plantar pressure measurement of non-pathological subjects using a dynamic pedobarograph. Gait & Posture 1997; 6: 50–55. 20. Rietdyk S, Patla A E Does the step length requirement in the subsequent step influence the strategies used for step length regulation in the current step? Human Movement Science 1994; 13: 109–127.
The Foot (1999) 9, 145–149