- Email: [email protected]
Simultaneous measurement of plantar pressure and shear forces in diabetic individuals
Gait and Posture 15 (2002) 101– 107 www.elsevier.com/locate/gaitpost
Simultaneous measurement of plantar pressure and shear forces in diabetic individuals Julie E. Perry a, James O. Hall b, Brian L. Davis a,* a
Department of Biomedical Engineering, Lerner Research Institute, The Cle6eland Clinic Foundation, ND20, 9500 Euclid A6enue, Cle6eland, OH44195, USA b Department of Orthopaedic Surgery, The Cle6eland Clinic Foundation, 9500 Euclid A6enue, Cle6eland, OH44195, USA Accepted 11 July 2001
Abstract Plantar foot ulceration is a diabetic complication whose underlying causative factors are still not fully understood. The goal of the current work was to simultaneously record plantar pressure and shear and examine the interrelationship of these forces; specifically, if peak shear and pressure occurred at the same site/time and whether adjacent shear forces had a greater tendency to be directed towards or away from each other. A custom built 16 transducer array was used to record forefoot shear and pressure during gait initiation in a cohort of 12 neuropathic diabetic individuals. The individuals were barefoot and the transducers were covered with a 5 mm thick layer of Minorplast. The greatest pressure occurred in the medial metatarsal heads (189 kPa) and the greatest shear in the lateral metatarsal heads (33 kPa). The interaction of the shear forces revealed that the plantar tissue was stretched to a greater magnitude than it was bunched (24 kPa vs 12 kPa, averaged over all regions). Normal distributions were determined for stretching and bunching in both the medial– lateral and anterior– posterior directions. When shear and pressure were considered in combination, half of the neuropathic individuals had peak shear and pressure occurring at the same site. These peak stresses did not occur at the same time (average difference of 0.186 s). The results of this study help to further characterize tissue stresses experienced on the plantar surface of the foot during gait initiation in neuropathic diabetic individuals. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Diabetes; Foot; Plantar pressure; Shear stress
1. Introduction For individuals with peripheral sensory neuropathy secondary to diabetes mellitus, the presence of ulcers on the plantar surface of the foot can be devastating [1]. Without proper care ulceration can be followed by infection, gangrene and amputation. In fact, a foot ulcer is the initiating factor in 85% of all diabetes related amputations [2]. Among the diabetic population the rate of amputation is 15 times greater than in the general population [3] and associated with this high rate of amputation are medical costs in excess of two billion U.S. dollars a year [4].
* Corresponding author. Tel.: + 1-216-444-1055; fax: +1-216-4449198. E-mail address: [email protected] (B.L. Davis).
One of the primary means by which to reduce the incidence of ulceration is through the early identification of individuals at risk of developing foot lesions. To this end, studies have examined the relationship between ulceration and areas of high plantar pressure [5–7] or shear force [8]. While it has been shown that regions of either high pressure or shear are associated with ulceration, diabetic individuals may ulcerate at pressure levels which are not commonly considered to be harmful [9]. This would suggest that shear and pressure need to be considered together since it may be that certain combinations of vertical and shear stresses result in loading scenarios that are more harmful than others [10]. For example, it has been demonstrated by others in the context of skin ulceration in swine [11] and blood occlusion in sitting paraplegic patients [12] that the presence of shear stresses dramatically reduces the pressure necessary to bring about adverse events.
0966-6362/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 6 - 6 3 6 2 ( 0 1 ) 0 0 1 7 6 - X
102
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
The simultaneous measurement of shear and pressure is technically demanding and until recently researchers had not developed the necessary instrumentation. In instances where the instrumentation has been developed the investigators have either chosen not to measure all three forces simultaneously [8,13] or have reported only on the design and construction of the device [14]. The purpose of this investigation was to use instrumentation developed in our laboratory [15] to simultaneously record the distribution of forefoot shear and pressure during the initiation of gait in order to more fully characterize tissue stresses on the plantar surface of the neuropathic diabetic foot. Specifically, we were interested in determining: (a) if peak shear and pressure occurred at the same site and time; and (b) whether adjacent shear forces had a greater tendency to be directed towards each other (resulting in bunching of the tissue) or away from each other (resulting in stretching of the tissue).
2. Methods
2.1. Subjects Twelve individuals who had diabetes and loss of protective sensation were recruited from a database of patients receiving podiatric care at The Cleveland Clinic Foundation. The subjects had no gross structural deformities of the foot, such as a Charcot fracture, but mild clawing of the toes was permissible. The presence of neuropathy (loss of sensation) in these diabetic individuals was identified based on Semmes-Weinstein monofilament (Fred Sammons Inc., Brookfield, Illinois) testing of the plantar surface of the hallux on both feet. Testing was carried out with monofilament grades of 4.17, 5.07 and 6.10 using a method of modified forcedchoice or interval comparison [16]. A monofilament grade of 5.07 was used to determine the presence/absence of protective sensation [16,17]. Individuals not capable of feeling the 5.07 filament were defined as neuropathic. Among the 12 individuals, only one could perceive the 6.10 filament while the other 11 could not feel the 6.10 filament (Table 1). Although inclusion in the study was based on monofilament testing, a Vibratron II (Physitemp Instruments Inc., Clifton, New Jersey) was also used to test vibratory perception under the hallux. Values for this test range from 0.1 (very good sensation) to 21.5 (extremely poor sensation). Nine of the neuropathic individuals could not feel the highest level of vibration, and the remaining three individuals had values that were greater than 18.8 (Table 1). Based on the monofilament and vibration tests, the subject group for this study was clearly defined as neuropathic.
2.2. Data collection Prior to data collection the study was fully explained to the subjects and written consent was obtained in accordance with institutional policy. The subjects were first seen by a podiatrist for routine removal of callus because callus has been shown to result in elevated plantar pressures [18]. An outline of the subject’s foot was then traced onto a schematic of the transducer array. The purpose of this tracing was to assist in maintaining a consistent alignment of the foot on the array for all data collection trials and to permit the orientation of the shear and pressure values to regions of the foot as defined below. Plantar shear and pressure data were collected simultaneously using a custom-built system consisting of 16 transducers, each with a surface area measuring 2.5× 2.5 cm, arranged in a 4× 4 array. Spacing between transducers was 1.5 mm. The transducer array was set flush in a 92 ×122 cm platform. Further details about the design and validation of the device have been published previously [15]. Data were sampled at a rate of 37 frames/s. Shear and pressure data were collected on the right forefoot during the initiation of gait. Subjects were tested in their bare feet and stood with their right forefoot on the transducer array. They initiated walking with the left foot and took a series of 2–3 steps. Three trials were collected. The device was designed such that the effect of different surface materials on shear and pressure could be tested by attaching individual caps to the top of each transducer. In this study, 5 mm thick Minorplast (P.W. Minor, Batavia, NY), a soft pliable material commonly used as an insole material in therapeutic shoes, was chosen for investigation. Table 1 Characteristics of the neuropathic diabetic subjects No. of subjects Age (years) Height (cm) Weight (kg) Weight index (kg/cm) Duration of diabetes (years) Vibration–perceptiona (vibration units) Monofilament gradeb 4.17 5.07 6.10 \6.10
12 (6 men) 59 9 12 (38–77) 170 910 (155–194) 87 925 (57–146) 0.51 90.11 (0.35–0.76) 21 910 (5–41) 21.5 91.1 (18.8–22.0) – – 1 11
Values are presented as the mean 9standard deviation, with the range in parentheses. a The ceiling value for this test is 21.5, nine individuals could not feel this level of vibration so they were assigned a value of 22.0. b Monofilament grade number =Log10 of bending force (measured in g).
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
103
Fig. 1. Representation of shear and pressure transducer array. (a) Depiction of shear forces and how the components of the force vectors were used to determine whether the tissue was being bunched (B) or stretched (S). See text for details. (b) Location of each transducer in the array and division of the foot into the three regions of interest. MTHs, metatarsal heads.
2.3. Data analysis For each of the 16 posts the maximum pressure and shear values were determined. In addition to the magnitude of the shear stresses the direction of the stresses was recorded. The medial– lateral and anterior– posterior components of the applied shear stresses at adjacent transducers were used to calculate bunching (adjacent forces acting towards each other) and stretching (adjacent forces acting away from each other) indices for each transducer in each of the four directions —medial, lateral, anterior and posterior (Fig. 1a). The values from the three trials were averaged to obtain the mean peak pressure, shear, bunching and stretching indices for each transducer. The forefoot was then divided into three regions: the toes, the medial metatarsal heads (MTHs), and the lateral MTHs (Fig. 1b). The first and second MTHs comprised the medial MTH region and the third, fourth and fifth heads comprised the lateral MTH region. In each of the three regions the overall mean peak value for each of the four variables was determined. The peak values for each of the variables would not necessarily be from the same transducer in each region. Additionally, a variable called ‘corresponding shear’ was determined for each region. This value was defined as the mean peak shear value recorded for the transducer that exhibited the mean peak pressure value for each particular region of the foot. Statistical analysis consisted of analysis of variance and, where appropriate, post-hoc multiple comparison tests using Student’s t-tests. A value of P B 0.05 was considered significant. Analyses were carried out to determine differences among the three foot regions. In addition to the above analyses, a more thorough examination of the stretching and bunching indices was undertaken because these measures have not been previously discussed in the literature. For each region of the foot, the bunching and stretching indices were
examined in both the medial–lateral direction and the anterior–posterior direction. Histograms were created to represent the distribution of these measures in the study population and then normal distribution curves were determined to more fully characterize these indices. The normal distribution function was of the form: f(y)= (2y| 2) − 1/2 exp[− (y− v)2/2| 2] where y is the response variable, v is the population mean, and | is the population standard deviation [19].
3. Results The neuropathic individuals experienced the highest plantar pressures in the region of the medial MTHs and the lowest pressures in the region of the toes (Fig. 2a). The pressure in the toe region (80 kPa9 36) was significantly less (PB 0.001) than that in both the medial MTHs (189 kPa 9 64) and the lateral MTHs (171 kPa 9 59). The greatest shear stresses occurred in the lateral MTHs (Fig. 2b). The magnitude of the shear stresses in the lateral MTHs (33 kPa9 9) was almost twice that of the toes (18 kPa97) and the medial MTHs (18 kPa9 6). This regional difference was significant (PB0.001). Corresponding shear, the peak shear value recorded for the same transducer that exhibited the peak pressure, showed the same patterns as maximum shear (Fig. 2b). The greatest corresponding shear occurred in the lateral MTHs (32 kPa 9 9) and was significantly (PB0.001) greater than that in the toes (16 kPa9 7) and medial MTHs (13 kPa 9 6). The interaction of the shear stresses at adjacent sites revealed that for all regions of the forefoot the tissue was stretched to a greater degree than it was bunched (Fig. 3). Averaged across all regions the magnitude of the stretching index (249 9 kPa) was twice as large as the bunching index (129 8 kPa). The only significant (P= 0.029) difference among regions was a smaller
104
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
Fig. 2. Mean peak (a) pressure and (b) shear and corresponding shear in each of the three regions of the foot for the neuropathic diabetic individuals. MTHs, metatarsal heads. Refer to text for definition of corresponding shear. * denotes statistical difference with respect to foot region (see text for P values).
bunching index for the medial MTHs (7 kPa 9 4) relative to the other two regions (toes, 15 kPa9 10; lateral MTHs, 15 kPa 9 7). The stretching and bunching indices were then examined in further detail by looking at the medial– lateral and anterior–posterior components of the indices. For both stretching and bunching there were significant regional differences in the medial– lateral direction but not in the anterior– posterior direction (Fig. 4). For the stretching index there was significantly less (P B 0.001)
Fig. 3. Mean peak indices for tissue stretching and bunching in each of the three regions of the foot for the neuropathic diabetic individuals. MTHs, metatarsal heads. Refer to text for definition of stretching and bunching indices. * denotes statistical difference with respect to foot region (see text for P values).
Fig. 4. Mean peak values for the medial – lateral (M/L) and anterior – posterior (A/P) components of the (a) stretch and (b) bunch indices in each of the three regions of the foot for the neuropathic diabetic individuals. MTHs, metatarsal heads. * denotes statistical difference with respect to foot region (see text for P values).
medial–lateral stretching in the toes (10 kPa9 8) than in the medial MTHs (27 kPa9 6) and lateral MTHs (26 kPa 9 8). With regard to the bunching index, the medial MTHs had significantly less (P= 0.014) medial– lateral bunching (4 kPa9 3) than the toes (8 kPa96) and lateral MTHs (12 kPa9 7). Histograms of the medial–lateral/anterior –posterior stretching and bunching indices are overlaid with normal distribution curves in Fig. 5 to provide a more thorough description of the distribution of these forces on the plantar surface of the foot. The combined effects of shear and pressure were investigated by determining the number of instances in which the sites of maximum shear and pressure corresponded, i.e. were recorded from the same individual transducer within a region. For the twelve individuals, the greatest number of coincident occurrences was noted in the lateral MTHs (9), followed by the toes (6), and the medial MTHs (4). When the overall peak pressure and shear were considered irrespective of region, half (6) of the individuals had maximum shear and pressure occurring at the same site under the foot. For the six individuals who had overall peak shear and pressure occurring at the same site we decided to look at their data in more depth. Since all of the above results were based on the average of three walking trials
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
we chose to look at each of the trials individually. For each trial we determined if the sites of overall maximum shear and pressure corresponded and we looked at the timing of the maximum values. Shear and pressure were considered to have occurred at the same time if the difference in timing was 09 0.054 s. This value represents twice the inverse of the sampling frequency (37 Hz) of the measurement device. Of the 18 trials (6 subjects × 3 trials) half had peak shear and pressure occurring at the same site and of these, only one trial had the peak values occurring at the same instant in time (Table 2). The average difference in time was 0.18690.139 s, indicating that peak pressure occurred prior to peak shear. 4. Discussion Previous work by others, as outlined in the introduction, has focused on shear or pressure but has not been able to simultaneously characterize all three components (pressure, anterior– posterior shear, and medial– lateral shear). The current work has permitted us to investigate whether maximum shear and pressure occur at the same sites under the foot and to look at the interaction and distribution of these forces in more detail.
105
The magnitudes of the average peak shear forces in our study (18–33 kPa depending on region of foot) are comparable to those reported by others. Lord and colleagues [13] measured values of 19–61 kPa on four subjects while Tappin and Robertson [20] reported that shear forces were greater than 500 g/cm2 (approximately 50 kPa), although details were not provided. Both of these studies utilized individual 16 mm transducer discs placed at specific sites of interest on the plantar surface of the foot. Shaw and colleagues [21] noted that the magnitude of the shear forces was much less than that of the vertical forces. Expressed as a percent bodyweight, the difference was approximately six fold (peak shear 19% vs peak pressure 116%). In the current study, although the data was not normalized to bodyweight, the relationship was similar (peak shear 33 kPa, peak pressure 189 kPa, over five fold difference). Commenting on the difference in magnitude between shear and pressure, Shaw et al. stated that ‘‘shear forces would seem unlikely to play a significant role in plantar ulceration, unless shear forces (resulting in tissue stretching) have different effects on tissue than do vertical forces (resulting in tissue compression)…when technologies are developed that can measure shear forces at different sites on the foot, this question can be addressed more accurately’’.
Fig. 5. Regional distribution of the medial – lateral (M/L) and anterior – posterior (A/P) stretching and bunching indices for the neuropathic diabetic individuals. Histograms represent the distribution in the study population (N =12 subjects) and the lines represent normal distribution curves (see text) based on the histogram data. For the histogram plots the location of the bars represents a range and not a finite number, for example, in graph (a) the first black bar denotes three subjects had a M/L stretching index that was in the range of 0 – 5 kPa (not a value of 3 kPa).
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
106
Table 2 Breakdown of data for the 6 neuropathic individuals who had overall peak shear and pressure occurring at the same site Time differencea (s)
Subject
Trial
Peak shear (kPa)
Peak pressure (kPa)
Shear post
Pressure post
1
1 2 3
27 38 32
171 182 158
11 11 11
11 10 11
0.162 0.216 0.162
2
1 2 3
13 13 16
51 52 58
3 7 7
7 11 7
0.162 0.162 0.189
3
1 2 3
20 16 16
84 142 85
15 11 15
15 15 15
0.108 −0.027 0.027
4
1 2 3
30 36 29
126 112 119
11 11 15
11 11 10
0.108 0.108 0.135
5
1 2 3
22 18 23
85 81 102
7 11 11
11 10 11
0.243 0.000 0.324
6
1 2 3
17 37 21
147 159 181
11 15 11
15 15 15
0.027 0.486 −0.027
Data are presented for each of the three trials for each individual. Post number refers to location in array Fig. 1 and indicates the site experiencing the peak shear or pressure. Boldface numbers in post columns indicate peak shear and pressure occurring at the same site. Boldface number in time difference column denotes peak shear and pressure occurring at the same site and the same time. a Difference is defined as (time of peak shear−time of peak pressure). A value of 0 90.054 s is considered as no difference in timing, i.e. peak shear and pressure occurred at the same time (see text for details).
Our work has provided an initial insight into this area of study. Not only were we able to measure shear forces at different sites on the foot, but the fact that our system was a matrix of sensors permitted us to examine the action of shear forces without limiting ourselves by a priori placement of shear sensors as noted above in the work of Lord et al. [13] and Tappin and Robertson [20]. We have shown that the interaction of the shear forces resulted in greater stretching than bunching of the tissue. The exact role that this may play in the formation of ulcers on the foot is not clear at this time. When considering shear and pressure together, half of the neuropathic individuals experienced maximum shear and pressure at the same site. When each of their trials was examined independently, 50% had these peak forces occurring at the same site although in no individual did this occur for all three trials (Table 2). We were only looking at a total of three steps but the cumulative effect of this interaction over a period of walking may be enough to initiate skin breakdown. Although peak shear and pressure occurred at the same site on the foot they did not occur at the same time. This is in contrast to the results of Tappin and Robertson [20]. However, the time interval which defines these two forces as occurring at the ‘same time’ differs between the two studies. We chose to base our definition (2×1/sampling frequency) on a function of the measuring capability of the device. Tappin and Robertson defined time as a
percentage of stance time and used a measure based on standard error. Additionally, it must be noted that Tappin and Robertson were combining shear data from one study with pressure data from another. Our study does have certain limitations. We had to restrict our focus to the initiation phase of gait. In order to orient the foot to the array of transducers for purposes of analysis it was necessary to start with the foot positioned on the array. This necessarily limits analysis to the push-off phase of the initial step. We could not do a standard first-step analysis [22] in which the subject would have stepped on and then off of the device. The size of the current transducers must also be kept in mind when evaluating the results of the study. With dimensions of 2.5× 2.5 cm the transducers are relatively large and as reported by Davis et al. [23] this may result in underestimation of the actual stresses by up to 15%. The large size also limits the ability to discriminate between two closely adjacent areas of high stresses. In summary, the present work has shown that, in our cohort of neuropathic diabetic individuals, peak shear and pressure occurred at the same site on the foot but not at the same time. The primary advantage of our system is the ability to record shear and pressure simultaneously, which eliminates the problem of having to combine data from different studies or even different steps. Additionally, since the device is a matrix of
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
sensors our evaluation is not limited to certain sites on the foot that are determined prior to testing, as is the case when individual transducers must be used. Both of these features, simultaneous recording of shear and pressure and an array of transducers, allow us to obtain a more robust and meaningful picture of what is occurring on the plantar surface of the foot. Plantar pressure measurements are well documented in the literature but shear measurement and analysis are still in their infancy. To our knowledge, reports of the simultaneous measurement of shear and pressure, the distribution of shear stresses under the forefoot, and the interaction of these shear stresses to create tissue bunching or stretching are very limited or non-existent. The current work provides a more comprehensive look at the stresses acting on the neuropathic diabetic foot. This ability to look at the interrelationship of shear and pressure will help to further our understanding of the factors that play a role in diabetic plantar ulceration.
Acknowledgements This work was supported by a grant from the Juvenile Diabetes Foundation International. The authors wish to thank Amrik Shah, Ph.D., Department of Biostatistics and Epidemiology at The Cleveland Clinic Foundation, for providing statistical support.
References [1] American Diabetes Association. Foot care in patients with diabetes mellitus [position statement]. Diabetes Care 1997;20(S1):S31 – 2. [2] Apelqvist J, Larsson J. What is the most effective way to reduce incidence of amputation in the diabetic foot? Diabetes/ Metabolism Research and Reviews 2000;16(S1):S75 –83. [3] Levy LA. Epidemiology and prevention of diabetic foot disease. In: Frykberg RG, editor. The High Risk Foot in Diabetes Mellitus. New York: Churchill Livingstone, 1991:23 – 30. [4] Reiber GE. Diabetic foot care. Financial implications and practice guidelines. Diabetes Care 1992;15(S1):29 – 31. [5] Boulton AJM, Hardisty CA, Betts RP, Franks CI, Worth RC, Ward JD, Duckworth T. Dynamic foot pressure and other studies as diagnostic and management aids in diabetic neuropathy. Diabetes Care 1983;6:26 –33. [6] Ctercteko GC, Dhanendran M, Hutton WC, LeQuesne LP. Vertical forces acting on the feet of diabetic patients with neuropathic ulceration. British Journal of Surgery 1981;68:608 – 14.
107
[7] Stokes IAF, Faris IB, Hutton WC. The neuropathic ulcer and loads on the foot in diabetic patients. Acta Orthopaedic Scandinavica 1975;46:839 – 47. [8] Pollard JP, LeQuesne LP. Method of healing diabetic forefoot ulcers. British Medical Journal 1983;286:436 – 7. [9] Hsi WL, Ulbrecht JS, Perry JE, Norkitis AJ, Cavanagh PR. Plantar pressure threshold for ulceration risk using the EMED SF platform [abstract]. Diabetes 1993;42(S1):103a. [10] Davis BL. Foot ulceration: hypotheses concerning shear and vertical forces acting on adjacent regions of skin. Medical Hypotheses 1993;40:44 – 7. [11] Dinsdale SM. Decubitus ulcers: role of pressure and friction in causation. Archives of Physical Medicine and Rehabilitation 1974;55:147 – 52. [12] Bennett L, Kavner D, Lee BY, Trainor FS, Lewis JM. Skin stress and blood flow in sitting paraplegic patients. Archives of Physical Medicine and Rehabilitation 1984;65:186 – 90. [13] Lord M, Hosein R, Williams RB. Method of in-shoe shear stress measurement. Journal of Biomedical Engineering 1992;14:181 – 6. [14] Huo M, Nicol K. 3-D force distribution measuring system [abstract]. In: Hakkinen K, Keskinen KL, Komi PV, Mero A, editors. XVth Congress of the International Society of Biomechanics: Book of Abstracts. Jyvaskyla, Finland: Gummerus Printing, 1995:410 – 1. [15] Davis BL, Perry JE, Neth DC, Waters KC. A device for simultaneous measurement of pressure and shear force distribution on the plantar surface of the foot. Journal of Applied Biomechanics 1997;14:93 – 104. [16] Holewski JJ, Stess RM, Graf PM, Grunfeld C. Aesthesiometry: quantification of cutaneous pressure sensation in diabetic peripheral neuropathy. Journal of Rehabilitation Research and Development 1988;25:1 – 10. [17] Birke JA, Sims DS. Plantar sensory threshold in the ulcerative foot. Leprosy Review 1986;57:261 – 7. [18] Young MJ, Cavanagh PR, Thomas G, Johnson MM, Murray H, Boulton AJ. The effect of callus removal on dynamic plantar foot pressures in diabetic patients. Diabetic Medicine 1992;9:55 – 7. [19] Mason RL, Gunst RF, Hess JL. Statistical Design and Analysis of Experiments: With Applications to Engineering and Science. New York: John Wiley & Sons, Inc., 1989:240. [20] Tappin JS, Robertson KP. Study of the relative timing of shear forces on the sole of the forefoot during walking. Journal of Biomedical Engineering 1991;13:39 – 42. [21] Shaw JE, van Schie CHM, Carrington AL, Abbott C, Boulton AJM. An analysis of dynamic forces transmitted through the foot in diabetic neuropathy. Diabetes Care 1998;21:1955 –9. [22] Cavanagh PR, Ulbrecht JS. Biomechanics of the diabetic foot: a quantitative approach to the assessment of neuropathy, deformity, and plantar pressure. In: Jahss MH, editor. Disorders of the Foot and Ankle, 2nd. Philadelphia: W.B Saunders, 1991:1864 – 907 [Chapter 66]. [23] Davis BL, Cothren RM, Quesada P, Hanson SB, Perry JE. Frequency content of normal and diabetic plantar pressure profiles: implications for the selection of transducer sizes. Journal of Biomechanics 1996;29:979 – 83.
Simultaneous measurement of plantar pressure and shear forces in diabetic individuals Julie E. Perry a, James O. Hall b, Brian L. Davis a,* a
Department of Biomedical Engineering, Lerner Research Institute, The Cle6eland Clinic Foundation, ND20, 9500 Euclid A6enue, Cle6eland, OH44195, USA b Department of Orthopaedic Surgery, The Cle6eland Clinic Foundation, 9500 Euclid A6enue, Cle6eland, OH44195, USA Accepted 11 July 2001
Abstract Plantar foot ulceration is a diabetic complication whose underlying causative factors are still not fully understood. The goal of the current work was to simultaneously record plantar pressure and shear and examine the interrelationship of these forces; specifically, if peak shear and pressure occurred at the same site/time and whether adjacent shear forces had a greater tendency to be directed towards or away from each other. A custom built 16 transducer array was used to record forefoot shear and pressure during gait initiation in a cohort of 12 neuropathic diabetic individuals. The individuals were barefoot and the transducers were covered with a 5 mm thick layer of Minorplast. The greatest pressure occurred in the medial metatarsal heads (189 kPa) and the greatest shear in the lateral metatarsal heads (33 kPa). The interaction of the shear forces revealed that the plantar tissue was stretched to a greater magnitude than it was bunched (24 kPa vs 12 kPa, averaged over all regions). Normal distributions were determined for stretching and bunching in both the medial– lateral and anterior– posterior directions. When shear and pressure were considered in combination, half of the neuropathic individuals had peak shear and pressure occurring at the same site. These peak stresses did not occur at the same time (average difference of 0.186 s). The results of this study help to further characterize tissue stresses experienced on the plantar surface of the foot during gait initiation in neuropathic diabetic individuals. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Diabetes; Foot; Plantar pressure; Shear stress
1. Introduction For individuals with peripheral sensory neuropathy secondary to diabetes mellitus, the presence of ulcers on the plantar surface of the foot can be devastating [1]. Without proper care ulceration can be followed by infection, gangrene and amputation. In fact, a foot ulcer is the initiating factor in 85% of all diabetes related amputations [2]. Among the diabetic population the rate of amputation is 15 times greater than in the general population [3] and associated with this high rate of amputation are medical costs in excess of two billion U.S. dollars a year [4].
* Corresponding author. Tel.: + 1-216-444-1055; fax: +1-216-4449198. E-mail address: [email protected] (B.L. Davis).
One of the primary means by which to reduce the incidence of ulceration is through the early identification of individuals at risk of developing foot lesions. To this end, studies have examined the relationship between ulceration and areas of high plantar pressure [5–7] or shear force [8]. While it has been shown that regions of either high pressure or shear are associated with ulceration, diabetic individuals may ulcerate at pressure levels which are not commonly considered to be harmful [9]. This would suggest that shear and pressure need to be considered together since it may be that certain combinations of vertical and shear stresses result in loading scenarios that are more harmful than others [10]. For example, it has been demonstrated by others in the context of skin ulceration in swine [11] and blood occlusion in sitting paraplegic patients [12] that the presence of shear stresses dramatically reduces the pressure necessary to bring about adverse events.
0966-6362/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 6 - 6 3 6 2 ( 0 1 ) 0 0 1 7 6 - X
102
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
The simultaneous measurement of shear and pressure is technically demanding and until recently researchers had not developed the necessary instrumentation. In instances where the instrumentation has been developed the investigators have either chosen not to measure all three forces simultaneously [8,13] or have reported only on the design and construction of the device [14]. The purpose of this investigation was to use instrumentation developed in our laboratory [15] to simultaneously record the distribution of forefoot shear and pressure during the initiation of gait in order to more fully characterize tissue stresses on the plantar surface of the neuropathic diabetic foot. Specifically, we were interested in determining: (a) if peak shear and pressure occurred at the same site and time; and (b) whether adjacent shear forces had a greater tendency to be directed towards each other (resulting in bunching of the tissue) or away from each other (resulting in stretching of the tissue).
2. Methods
2.1. Subjects Twelve individuals who had diabetes and loss of protective sensation were recruited from a database of patients receiving podiatric care at The Cleveland Clinic Foundation. The subjects had no gross structural deformities of the foot, such as a Charcot fracture, but mild clawing of the toes was permissible. The presence of neuropathy (loss of sensation) in these diabetic individuals was identified based on Semmes-Weinstein monofilament (Fred Sammons Inc., Brookfield, Illinois) testing of the plantar surface of the hallux on both feet. Testing was carried out with monofilament grades of 4.17, 5.07 and 6.10 using a method of modified forcedchoice or interval comparison [16]. A monofilament grade of 5.07 was used to determine the presence/absence of protective sensation [16,17]. Individuals not capable of feeling the 5.07 filament were defined as neuropathic. Among the 12 individuals, only one could perceive the 6.10 filament while the other 11 could not feel the 6.10 filament (Table 1). Although inclusion in the study was based on monofilament testing, a Vibratron II (Physitemp Instruments Inc., Clifton, New Jersey) was also used to test vibratory perception under the hallux. Values for this test range from 0.1 (very good sensation) to 21.5 (extremely poor sensation). Nine of the neuropathic individuals could not feel the highest level of vibration, and the remaining three individuals had values that were greater than 18.8 (Table 1). Based on the monofilament and vibration tests, the subject group for this study was clearly defined as neuropathic.
2.2. Data collection Prior to data collection the study was fully explained to the subjects and written consent was obtained in accordance with institutional policy. The subjects were first seen by a podiatrist for routine removal of callus because callus has been shown to result in elevated plantar pressures [18]. An outline of the subject’s foot was then traced onto a schematic of the transducer array. The purpose of this tracing was to assist in maintaining a consistent alignment of the foot on the array for all data collection trials and to permit the orientation of the shear and pressure values to regions of the foot as defined below. Plantar shear and pressure data were collected simultaneously using a custom-built system consisting of 16 transducers, each with a surface area measuring 2.5× 2.5 cm, arranged in a 4× 4 array. Spacing between transducers was 1.5 mm. The transducer array was set flush in a 92 ×122 cm platform. Further details about the design and validation of the device have been published previously [15]. Data were sampled at a rate of 37 frames/s. Shear and pressure data were collected on the right forefoot during the initiation of gait. Subjects were tested in their bare feet and stood with their right forefoot on the transducer array. They initiated walking with the left foot and took a series of 2–3 steps. Three trials were collected. The device was designed such that the effect of different surface materials on shear and pressure could be tested by attaching individual caps to the top of each transducer. In this study, 5 mm thick Minorplast (P.W. Minor, Batavia, NY), a soft pliable material commonly used as an insole material in therapeutic shoes, was chosen for investigation. Table 1 Characteristics of the neuropathic diabetic subjects No. of subjects Age (years) Height (cm) Weight (kg) Weight index (kg/cm) Duration of diabetes (years) Vibration–perceptiona (vibration units) Monofilament gradeb 4.17 5.07 6.10 \6.10
12 (6 men) 59 9 12 (38–77) 170 910 (155–194) 87 925 (57–146) 0.51 90.11 (0.35–0.76) 21 910 (5–41) 21.5 91.1 (18.8–22.0) – – 1 11
Values are presented as the mean 9standard deviation, with the range in parentheses. a The ceiling value for this test is 21.5, nine individuals could not feel this level of vibration so they were assigned a value of 22.0. b Monofilament grade number =Log10 of bending force (measured in g).
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
103
Fig. 1. Representation of shear and pressure transducer array. (a) Depiction of shear forces and how the components of the force vectors were used to determine whether the tissue was being bunched (B) or stretched (S). See text for details. (b) Location of each transducer in the array and division of the foot into the three regions of interest. MTHs, metatarsal heads.
2.3. Data analysis For each of the 16 posts the maximum pressure and shear values were determined. In addition to the magnitude of the shear stresses the direction of the stresses was recorded. The medial– lateral and anterior– posterior components of the applied shear stresses at adjacent transducers were used to calculate bunching (adjacent forces acting towards each other) and stretching (adjacent forces acting away from each other) indices for each transducer in each of the four directions —medial, lateral, anterior and posterior (Fig. 1a). The values from the three trials were averaged to obtain the mean peak pressure, shear, bunching and stretching indices for each transducer. The forefoot was then divided into three regions: the toes, the medial metatarsal heads (MTHs), and the lateral MTHs (Fig. 1b). The first and second MTHs comprised the medial MTH region and the third, fourth and fifth heads comprised the lateral MTH region. In each of the three regions the overall mean peak value for each of the four variables was determined. The peak values for each of the variables would not necessarily be from the same transducer in each region. Additionally, a variable called ‘corresponding shear’ was determined for each region. This value was defined as the mean peak shear value recorded for the transducer that exhibited the mean peak pressure value for each particular region of the foot. Statistical analysis consisted of analysis of variance and, where appropriate, post-hoc multiple comparison tests using Student’s t-tests. A value of P B 0.05 was considered significant. Analyses were carried out to determine differences among the three foot regions. In addition to the above analyses, a more thorough examination of the stretching and bunching indices was undertaken because these measures have not been previously discussed in the literature. For each region of the foot, the bunching and stretching indices were
examined in both the medial–lateral direction and the anterior–posterior direction. Histograms were created to represent the distribution of these measures in the study population and then normal distribution curves were determined to more fully characterize these indices. The normal distribution function was of the form: f(y)= (2y| 2) − 1/2 exp[− (y− v)2/2| 2] where y is the response variable, v is the population mean, and | is the population standard deviation [19].
3. Results The neuropathic individuals experienced the highest plantar pressures in the region of the medial MTHs and the lowest pressures in the region of the toes (Fig. 2a). The pressure in the toe region (80 kPa9 36) was significantly less (PB 0.001) than that in both the medial MTHs (189 kPa 9 64) and the lateral MTHs (171 kPa 9 59). The greatest shear stresses occurred in the lateral MTHs (Fig. 2b). The magnitude of the shear stresses in the lateral MTHs (33 kPa9 9) was almost twice that of the toes (18 kPa97) and the medial MTHs (18 kPa9 6). This regional difference was significant (PB0.001). Corresponding shear, the peak shear value recorded for the same transducer that exhibited the peak pressure, showed the same patterns as maximum shear (Fig. 2b). The greatest corresponding shear occurred in the lateral MTHs (32 kPa 9 9) and was significantly (PB0.001) greater than that in the toes (16 kPa9 7) and medial MTHs (13 kPa 9 6). The interaction of the shear stresses at adjacent sites revealed that for all regions of the forefoot the tissue was stretched to a greater degree than it was bunched (Fig. 3). Averaged across all regions the magnitude of the stretching index (249 9 kPa) was twice as large as the bunching index (129 8 kPa). The only significant (P= 0.029) difference among regions was a smaller
104
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
Fig. 2. Mean peak (a) pressure and (b) shear and corresponding shear in each of the three regions of the foot for the neuropathic diabetic individuals. MTHs, metatarsal heads. Refer to text for definition of corresponding shear. * denotes statistical difference with respect to foot region (see text for P values).
bunching index for the medial MTHs (7 kPa 9 4) relative to the other two regions (toes, 15 kPa9 10; lateral MTHs, 15 kPa 9 7). The stretching and bunching indices were then examined in further detail by looking at the medial– lateral and anterior–posterior components of the indices. For both stretching and bunching there were significant regional differences in the medial– lateral direction but not in the anterior– posterior direction (Fig. 4). For the stretching index there was significantly less (P B 0.001)
Fig. 3. Mean peak indices for tissue stretching and bunching in each of the three regions of the foot for the neuropathic diabetic individuals. MTHs, metatarsal heads. Refer to text for definition of stretching and bunching indices. * denotes statistical difference with respect to foot region (see text for P values).
Fig. 4. Mean peak values for the medial – lateral (M/L) and anterior – posterior (A/P) components of the (a) stretch and (b) bunch indices in each of the three regions of the foot for the neuropathic diabetic individuals. MTHs, metatarsal heads. * denotes statistical difference with respect to foot region (see text for P values).
medial–lateral stretching in the toes (10 kPa9 8) than in the medial MTHs (27 kPa9 6) and lateral MTHs (26 kPa 9 8). With regard to the bunching index, the medial MTHs had significantly less (P= 0.014) medial– lateral bunching (4 kPa9 3) than the toes (8 kPa96) and lateral MTHs (12 kPa9 7). Histograms of the medial–lateral/anterior –posterior stretching and bunching indices are overlaid with normal distribution curves in Fig. 5 to provide a more thorough description of the distribution of these forces on the plantar surface of the foot. The combined effects of shear and pressure were investigated by determining the number of instances in which the sites of maximum shear and pressure corresponded, i.e. were recorded from the same individual transducer within a region. For the twelve individuals, the greatest number of coincident occurrences was noted in the lateral MTHs (9), followed by the toes (6), and the medial MTHs (4). When the overall peak pressure and shear were considered irrespective of region, half (6) of the individuals had maximum shear and pressure occurring at the same site under the foot. For the six individuals who had overall peak shear and pressure occurring at the same site we decided to look at their data in more depth. Since all of the above results were based on the average of three walking trials
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
we chose to look at each of the trials individually. For each trial we determined if the sites of overall maximum shear and pressure corresponded and we looked at the timing of the maximum values. Shear and pressure were considered to have occurred at the same time if the difference in timing was 09 0.054 s. This value represents twice the inverse of the sampling frequency (37 Hz) of the measurement device. Of the 18 trials (6 subjects × 3 trials) half had peak shear and pressure occurring at the same site and of these, only one trial had the peak values occurring at the same instant in time (Table 2). The average difference in time was 0.18690.139 s, indicating that peak pressure occurred prior to peak shear. 4. Discussion Previous work by others, as outlined in the introduction, has focused on shear or pressure but has not been able to simultaneously characterize all three components (pressure, anterior– posterior shear, and medial– lateral shear). The current work has permitted us to investigate whether maximum shear and pressure occur at the same sites under the foot and to look at the interaction and distribution of these forces in more detail.
105
The magnitudes of the average peak shear forces in our study (18–33 kPa depending on region of foot) are comparable to those reported by others. Lord and colleagues [13] measured values of 19–61 kPa on four subjects while Tappin and Robertson [20] reported that shear forces were greater than 500 g/cm2 (approximately 50 kPa), although details were not provided. Both of these studies utilized individual 16 mm transducer discs placed at specific sites of interest on the plantar surface of the foot. Shaw and colleagues [21] noted that the magnitude of the shear forces was much less than that of the vertical forces. Expressed as a percent bodyweight, the difference was approximately six fold (peak shear 19% vs peak pressure 116%). In the current study, although the data was not normalized to bodyweight, the relationship was similar (peak shear 33 kPa, peak pressure 189 kPa, over five fold difference). Commenting on the difference in magnitude between shear and pressure, Shaw et al. stated that ‘‘shear forces would seem unlikely to play a significant role in plantar ulceration, unless shear forces (resulting in tissue stretching) have different effects on tissue than do vertical forces (resulting in tissue compression)…when technologies are developed that can measure shear forces at different sites on the foot, this question can be addressed more accurately’’.
Fig. 5. Regional distribution of the medial – lateral (M/L) and anterior – posterior (A/P) stretching and bunching indices for the neuropathic diabetic individuals. Histograms represent the distribution in the study population (N =12 subjects) and the lines represent normal distribution curves (see text) based on the histogram data. For the histogram plots the location of the bars represents a range and not a finite number, for example, in graph (a) the first black bar denotes three subjects had a M/L stretching index that was in the range of 0 – 5 kPa (not a value of 3 kPa).
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
106
Table 2 Breakdown of data for the 6 neuropathic individuals who had overall peak shear and pressure occurring at the same site Time differencea (s)
Subject
Trial
Peak shear (kPa)
Peak pressure (kPa)
Shear post
Pressure post
1
1 2 3
27 38 32
171 182 158
11 11 11
11 10 11
0.162 0.216 0.162
2
1 2 3
13 13 16
51 52 58
3 7 7
7 11 7
0.162 0.162 0.189
3
1 2 3
20 16 16
84 142 85
15 11 15
15 15 15
0.108 −0.027 0.027
4
1 2 3
30 36 29
126 112 119
11 11 15
11 11 10
0.108 0.108 0.135
5
1 2 3
22 18 23
85 81 102
7 11 11
11 10 11
0.243 0.000 0.324
6
1 2 3
17 37 21
147 159 181
11 15 11
15 15 15
0.027 0.486 −0.027
Data are presented for each of the three trials for each individual. Post number refers to location in array Fig. 1 and indicates the site experiencing the peak shear or pressure. Boldface numbers in post columns indicate peak shear and pressure occurring at the same site. Boldface number in time difference column denotes peak shear and pressure occurring at the same site and the same time. a Difference is defined as (time of peak shear−time of peak pressure). A value of 0 90.054 s is considered as no difference in timing, i.e. peak shear and pressure occurred at the same time (see text for details).
Our work has provided an initial insight into this area of study. Not only were we able to measure shear forces at different sites on the foot, but the fact that our system was a matrix of sensors permitted us to examine the action of shear forces without limiting ourselves by a priori placement of shear sensors as noted above in the work of Lord et al. [13] and Tappin and Robertson [20]. We have shown that the interaction of the shear forces resulted in greater stretching than bunching of the tissue. The exact role that this may play in the formation of ulcers on the foot is not clear at this time. When considering shear and pressure together, half of the neuropathic individuals experienced maximum shear and pressure at the same site. When each of their trials was examined independently, 50% had these peak forces occurring at the same site although in no individual did this occur for all three trials (Table 2). We were only looking at a total of three steps but the cumulative effect of this interaction over a period of walking may be enough to initiate skin breakdown. Although peak shear and pressure occurred at the same site on the foot they did not occur at the same time. This is in contrast to the results of Tappin and Robertson [20]. However, the time interval which defines these two forces as occurring at the ‘same time’ differs between the two studies. We chose to base our definition (2×1/sampling frequency) on a function of the measuring capability of the device. Tappin and Robertson defined time as a
percentage of stance time and used a measure based on standard error. Additionally, it must be noted that Tappin and Robertson were combining shear data from one study with pressure data from another. Our study does have certain limitations. We had to restrict our focus to the initiation phase of gait. In order to orient the foot to the array of transducers for purposes of analysis it was necessary to start with the foot positioned on the array. This necessarily limits analysis to the push-off phase of the initial step. We could not do a standard first-step analysis [22] in which the subject would have stepped on and then off of the device. The size of the current transducers must also be kept in mind when evaluating the results of the study. With dimensions of 2.5× 2.5 cm the transducers are relatively large and as reported by Davis et al. [23] this may result in underestimation of the actual stresses by up to 15%. The large size also limits the ability to discriminate between two closely adjacent areas of high stresses. In summary, the present work has shown that, in our cohort of neuropathic diabetic individuals, peak shear and pressure occurred at the same site on the foot but not at the same time. The primary advantage of our system is the ability to record shear and pressure simultaneously, which eliminates the problem of having to combine data from different studies or even different steps. Additionally, since the device is a matrix of
J.E. Perry et al. / Gait and Posture 15 (2002) 101–107
sensors our evaluation is not limited to certain sites on the foot that are determined prior to testing, as is the case when individual transducers must be used. Both of these features, simultaneous recording of shear and pressure and an array of transducers, allow us to obtain a more robust and meaningful picture of what is occurring on the plantar surface of the foot. Plantar pressure measurements are well documented in the literature but shear measurement and analysis are still in their infancy. To our knowledge, reports of the simultaneous measurement of shear and pressure, the distribution of shear stresses under the forefoot, and the interaction of these shear stresses to create tissue bunching or stretching are very limited or non-existent. The current work provides a more comprehensive look at the stresses acting on the neuropathic diabetic foot. This ability to look at the interrelationship of shear and pressure will help to further our understanding of the factors that play a role in diabetic plantar ulceration.
Acknowledgements This work was supported by a grant from the Juvenile Diabetes Foundation International. The authors wish to thank Amrik Shah, Ph.D., Department of Biostatistics and Epidemiology at The Cleveland Clinic Foundation, for providing statistical support.
References [1] American Diabetes Association. Foot care in patients with diabetes mellitus [position statement]. Diabetes Care 1997;20(S1):S31 – 2. [2] Apelqvist J, Larsson J. What is the most effective way to reduce incidence of amputation in the diabetic foot? Diabetes/ Metabolism Research and Reviews 2000;16(S1):S75 –83. [3] Levy LA. Epidemiology and prevention of diabetic foot disease. In: Frykberg RG, editor. The High Risk Foot in Diabetes Mellitus. New York: Churchill Livingstone, 1991:23 – 30. [4] Reiber GE. Diabetic foot care. Financial implications and practice guidelines. Diabetes Care 1992;15(S1):29 – 31. [5] Boulton AJM, Hardisty CA, Betts RP, Franks CI, Worth RC, Ward JD, Duckworth T. Dynamic foot pressure and other studies as diagnostic and management aids in diabetic neuropathy. Diabetes Care 1983;6:26 –33. [6] Ctercteko GC, Dhanendran M, Hutton WC, LeQuesne LP. Vertical forces acting on the feet of diabetic patients with neuropathic ulceration. British Journal of Surgery 1981;68:608 – 14.
107
[7] Stokes IAF, Faris IB, Hutton WC. The neuropathic ulcer and loads on the foot in diabetic patients. Acta Orthopaedic Scandinavica 1975;46:839 – 47. [8] Pollard JP, LeQuesne LP. Method of healing diabetic forefoot ulcers. British Medical Journal 1983;286:436 – 7. [9] Hsi WL, Ulbrecht JS, Perry JE, Norkitis AJ, Cavanagh PR. Plantar pressure threshold for ulceration risk using the EMED SF platform [abstract]. Diabetes 1993;42(S1):103a. [10] Davis BL. Foot ulceration: hypotheses concerning shear and vertical forces acting on adjacent regions of skin. Medical Hypotheses 1993;40:44 – 7. [11] Dinsdale SM. Decubitus ulcers: role of pressure and friction in causation. Archives of Physical Medicine and Rehabilitation 1974;55:147 – 52. [12] Bennett L, Kavner D, Lee BY, Trainor FS, Lewis JM. Skin stress and blood flow in sitting paraplegic patients. Archives of Physical Medicine and Rehabilitation 1984;65:186 – 90. [13] Lord M, Hosein R, Williams RB. Method of in-shoe shear stress measurement. Journal of Biomedical Engineering 1992;14:181 – 6. [14] Huo M, Nicol K. 3-D force distribution measuring system [abstract]. In: Hakkinen K, Keskinen KL, Komi PV, Mero A, editors. XVth Congress of the International Society of Biomechanics: Book of Abstracts. Jyvaskyla, Finland: Gummerus Printing, 1995:410 – 1. [15] Davis BL, Perry JE, Neth DC, Waters KC. A device for simultaneous measurement of pressure and shear force distribution on the plantar surface of the foot. Journal of Applied Biomechanics 1997;14:93 – 104. [16] Holewski JJ, Stess RM, Graf PM, Grunfeld C. Aesthesiometry: quantification of cutaneous pressure sensation in diabetic peripheral neuropathy. Journal of Rehabilitation Research and Development 1988;25:1 – 10. [17] Birke JA, Sims DS. Plantar sensory threshold in the ulcerative foot. Leprosy Review 1986;57:261 – 7. [18] Young MJ, Cavanagh PR, Thomas G, Johnson MM, Murray H, Boulton AJ. The effect of callus removal on dynamic plantar foot pressures in diabetic patients. Diabetic Medicine 1992;9:55 – 7. [19] Mason RL, Gunst RF, Hess JL. Statistical Design and Analysis of Experiments: With Applications to Engineering and Science. New York: John Wiley & Sons, Inc., 1989:240. [20] Tappin JS, Robertson KP. Study of the relative timing of shear forces on the sole of the forefoot during walking. Journal of Biomedical Engineering 1991;13:39 – 42. [21] Shaw JE, van Schie CHM, Carrington AL, Abbott C, Boulton AJM. An analysis of dynamic forces transmitted through the foot in diabetic neuropathy. Diabetes Care 1998;21:1955 –9. [22] Cavanagh PR, Ulbrecht JS. Biomechanics of the diabetic foot: a quantitative approach to the assessment of neuropathy, deformity, and plantar pressure. In: Jahss MH, editor. Disorders of the Foot and Ankle, 2nd. Philadelphia: W.B Saunders, 1991:1864 – 907 [Chapter 66]. [23] Davis BL, Cothren RM, Quesada P, Hanson SB, Perry JE. Frequency content of normal and diabetic plantar pressure profiles: implications for the selection of transducer sizes. Journal of Biomechanics 1996;29:979 – 83.