“Pressure Gradient” as an Indicator of Plantar Skin Injury

Subjects were recruited as part of a larger study seeking to develop a computational model of the diabetic foot. As the computational model was being developed, it became apparent that the change in pressures across the plantar foot at the location of the PPP, or the PPG, may be an important variable. The preliminary data from our computational model demonstrated potentially damaging shear stresses occurring inside the soft tissues where there was a large difference (or gradient) in pressure between two adjacent sites on the plantar foot.

Subjects were recruited from the Diabetic Foot Center, Volunteers for Health, the Diabetes Research Training Center at Washington University School of Medicine, and the BJC Health System in St. Louis, Missouri. Criteria for entry into the study were a history of diabetes, evidence of peripheral neuropathy (inability to sense the 5.07 Semmes Weinstein monofilament and vibratory perception threshold [VPT] >25 V), a palpable pedal pulse, and a history of a midfoot or forefoot plantar ulcer. Nonambulatory patients or patients with severe midfoot or rearfoot Charcot neuroarthropathy were excluded.

Sensation was tested using the 5.07 Semmes Weinstein monofilament and a Bio-Thesiometer (Bio-Medical Instrument, Newbury, OH). All subjects were unable to sense the 5.07 Semmes Weinstein monofilament on at least two sites on the plantar foot as described elsewhere (12). Sensation also was quantified with a Bio-Thesiometer. The head of the Bio-Thesiometer was held perpendicular to the plantar surface of the great toe, and the amplitude of the vibration was increased gradually. The VPT was defined as the lowest voltage the patient could perceive using a mean of three trials (2). A value of 50 V was assigned to those subjects unable to perceive the voltage even when turned to its maximum amplitude. The mean ± SD value for this group was 48.3 ± 4.1 V, indicating a severe level of neuropathy (Table 1). Ankle dorsiflexion range of motion was measured with a plastic, full-circle goniometer with the subject in a prone position, the knee extended, and the subtalar joint in a neutral position as described previously (12). All reported ulcers had occurred on the plantar surface of the midfoot or forefoot as described with other subject characteristics in Table 1. All subjects read and signed the informed medical consent according to the institutional review board approved protocol before entrance into the study.

Plantar pressures were recorded during walking using the F-Scan system (Tekscan, Boston, MA) and previously validated methods (3,13,14). All subjects wore Soletech shoes (Advance orthopedic footwear, style number E3010) that were fitted by a certified pedorthist or orthotist. The footwear was selected because it is used widely in clinical practices, meets all the guidelines of the “therapeutic shoe bill” (Medicare Part B), and can accommodate forefoot deformities or a custom orthotic device. For the purposes of this study, the shoe was worn with its standard insert and a thin cotton sock. Excessive callus that might affect pressure measures was trimmed before testing.

Each F-scan in-shoe sensor contains 960 sensing locations or pixels. Each pixel is 5.08 × 5.08 mm. A new sensor was trimmed to fit the shape of the subject’s foot and was placed between the subject’s sock and the insole. Subjects wore the sensor inside the shoe for several minutes and took about 40 steps before calibration for adequate sensor “break in” (3,13). The sensor was calibrated according to the manufacturer’s guidelines and was consistent with other investigators (3,13,14). Subjects were allowed to walk at their chosen walking speed, and data were collected at 50 Hz during two walking trials immediately after calibration. A mean of three steps chosen during the midportion of the walking trials was used for the PPP and PPG variables.

The PPP and PPG were determined using custom software. ASCII files containing all frames of data from three steps were processed by the software. The foot was divided in half from the distal toe to the proximal heel to determine the forefoot and rearfoot. The software selected the PPP at the rearfoot and forefoot. This value was checked with the PPP values obtained from the F-Scan software. A bicubic polynomial spline smoothing function was applied to the raw data before the PPG was determined to eliminate individual pixel outliers and to estimate pressure values at nodes located half the length between each sensor pixel. The reason for adding nodes between the sensor pixels was to increase the accuracy of the PPG calculation. The PPG then was determined in a defined area (3 × 3 F-Scan sensor pixels [231.3 mm²]) around the PPP by calculating the highest change in pressure (pressure gradient) from one node (half pixel apart) to the next according to rows and columns and by diagonal (Fig. 1). The pressure gradient values were calculated by subtracting the pressure in each node around the PPP from that in the adjacent node and dividing by the distance between the centers of the nodes.

The PPP and PPG in the forefoot and rearfoot were recorded for each subject. A PPP forefoot-to-rearfoot ratio was calculated by dividing the PPP in the forefoot by the PPP in the rearfoot (2). Likewise, a PPG forefoot-to-rearfoot ratio was calculated by dividing the PPG in the forefoot by the PPG in the rearfoot. Differences between PPP and PPG in the forefoot and in the rearfoot and the difference between the PPP and PPG forefoot-to-rearfoot ratio were determined using two-tailed paired t tests. The correlations between the PPP and PPG at the forefoot and rearfoot were determined using a Pearson product-moment correlation.

The mean PPG was 143% higher in the forefoot than in the rearfoot whereas the mean PPP was only 36% higher in the forefoot than in the rearfoot (Table 2) (P < 0.0001). Figure 2 illustrates an example of the PPP and the PPG at the forefoot and the rearfoot of a representative subject. The PPG forefoot-to-rearfoot ratio (2.84 ± 1.36) was nearly two times greater than the PPP forefoot-to-rearfoot ratio (1.48 ± 0.58) (P < 0.0001) (Table 2). The correlation between PPP and PPG was r = 0.59 at the forefoot and r = 0.75 at the rearfoot.

The primary results of this study indicate that the PPG is much higher in the forefoot than in the rearfoot (forefoot-to-rearfoot ratio), even when compared with the PPP forefoot-to-rearfoot ratio. These results may be important because the incidence of skin breakdown is much higher in the forefoot than in the rearfoot (2,11). Perhaps the PPG is an important indicator of tissue trauma because it indicates high levels of stress concentration in the soft tissues that may lead to skin breakdown.

The correlation between PPG and PPP was considerably higher in the rearfoot than in the forefoot (r = 0.75 vs. 0.59). PPP accounted for 57% of the variance of PPG at the heel but only 35% of the variance at the forefoot. There are several possible reasons that the PPG may be much higher in the forefoot than the rearfoot (even when compared with the PPP forefoot-to-rearfoot ratio) and that the PPG is more highly correlated to the PPP in the rearfoot than in the forefoot. First, the soft tissue thickness is 36–48% greater under the heel than that under the metatarsal heads (15). This increased soft tissue thickness probably helps to distribute plantar pressures to increased areas and attenuate peak pressures and pressure gradients (15,16). Furthermore, although the shape of the calcaneus tends to remain fairly intact after complications of peripheral neuropathy, the metatarsal heads are often deformed by an increased metatarsal phalangeal joint angle (hammer toe deformity) and arthropathy that can contribute to a high PPP (17) and PPG. Therefore, the structure of the calcaneus and the thickness of the soft tissue may provide a more consistent pressure gradient in the rearfoot than in the metatarsal heads in the forefoot. The primary difference between the PPG and the PPP is that the PPG represents the spatial change in the pressure in the region of the peak pressure. From a mechanical standpoint, a sharp change in high pressure, i.e., a high PPG, may lead to internal stress concentrations and shearing of soft tissues, causing injury. This higher change in peak pressure may be one reason that the forefoot experiences a higher incidence of skin breakdown than does the heel despite relatively similar PPPs during walking in shoes.

Another important reason for the higher incidence of skin breakdown in the forefoot than in the rearfoot is that the PPPs are higher in the forefoot than in the rearfoot during barefoot ambulation (2). Caselli et al. (2) studied the forefoot-to-rearfoot plantar pressure ratio in several patient groups including subjects with diabetes and severe neuropathy similar to that described in this study (VPT 48.3 ± 6.2 V). The forefoot PPP (6.2 kg/cm² [608 kPa]) and PPP forefoot-to-rearfoot ratio (2.3 ± 2.4) were much higher than the values reported in our study. Although the pressure sensor characteristics were similar to ones used in our study, the higher values reported in that study probably occurred because the subjects walked barefoot across a mat rather than walking in therapeutic footwear. PPPs in our study are similar to those reported by Pitei et al. (3), who reported forefoot and heel PPPs of 242 ± 25 and 240 ± 28 kPa, respectively. These values are slightly lower than the values reported in our study (Table 2), probably because Pitei et al. had subjects walk in therapeutic shoes that included multilayer inserts whereas subjects in this study walked in a standard insert.

We are aware of no other research studies that have specifically examined the PPG. The concept of a pressure gradient contributing to internal stress, however, is documented in the engineering literature (10). In addition, Prabhu et al. (18) described a parameter called the “power ratio,” which was used to accurately discriminate among groups of patients with neuropathy and to predict risk for skin breakdown. The power ratio is defined as the ratio of high-frequency power to total power in an image generated by a pedobarograph and appears to be influenced by plantar pressures that change rapidly across the surface of the foot (18). Hence, the power ratio and the PPG each appear to be measures of the change in pressure across the surface of the foot.

Mechanical stresses contributing to plantar skin breakdown appear to be a composite value of several indicators (19). The magnitude of stress clearly has been associated with the location of skin breakdown and is measured as the PPP (1,2,3,4), but the PPP alone cannot clearly predict skin breakdown (5). We believe the PPG is an indicator of plantar skin trauma that is related, but not identical, to the PPP, as reflected by the results of this study. The duration of pressure application to the skin also is thought to be important and is estimated by the pressure time integral (6). Direction of stress and the number of repetitions also are thought to be important (6,9). Although it was not the primary purpose of this study, we performed a post hoc analysis to determine whether there was an association of the location of the PPP, pressure time integral, and PPG to location of ulcer. In this small sample size, PPP, pressure time integral, and PPG occurred at the healed ulcer site in 56, 69, and 31% of the occurrences, respectively. One of the three variables was associated with the location of skin breakdown 81% of the time. We believe that all of these variables are related to one another but contain unique information to help predict trauma and skin breakdown.

This study is a first step in investigating the importance of the PPG as an indicator of plantar skin injury on the neuropathic foot. Clearly there are many mechanical and biological factors that contribute to skin breakdown (7). This study was a cross-sectional analysis of a single group of patients with diabetes, severe peripheral neuropathy, and a history of ulcers. Additional studies are required to determine the benefits and limitations of using PPG as an indicator of tissue injury and skin breakdown. Computational models could investigate the relationship between pressure gradients on the surface of the foot to stress concentrations within the soft tissues of the foot. PPG should be investigated in groups of subjects with various levels of neuropathy and histories of ulcers. Finally, prospective studies should be conducted to determine whether the PPG can be used with the PPP and the pressure time integral to help predict skin breakdown. Although additional studies are required, the results of this study indicate that the PPG may be an important variable to identify where stress concentrations that may lead to skin breakdown are located on the plantar foot.

Funding was provided by the National Center of Medical Rehabilitation and Research, National Institutes of Health (RO1 HD36895).

The authors acknowledge the Prevention and Control Research Core of the Washington University Diabetes Research and Training Center (P60 DK20579) for their assistance in subject recruitment.