Skip to main content
NCBI home page
VA Author Manuscripts logoLink to VA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 1.
Published in final edited form as: J Biomech. 2026 Feb 8;199:113201. doi: 10.1016/j.jbiomech.2026.113201

Quantifying differences in high-pressure region mapping between dynamic in-shoe and barefoot plantar pressure in diabetic subjects

Dylan Heino a,b, Scott Telfer a,b,c, Avocet Nagle-Christensen a, William R Ledoux a,b,c, Brittney C Muir a,b,*
PMCID: PMC13317507  NIHMSID: NIHMS2184137  PMID: 41775104

Abstract

Foot ulceration is a serious complication of diabetes, often linked to elevated forefoot pressures during walking. Custom accommodative insoles reduce peak plantar pressures, with greater reductions achieved when dynamic pressure data informs design. While in-shoe and barefoot data are used separately to design insoles, their agreement in pressure distribution and high-pressure region location has not been evaluated. This study compares pressure distribution and high-pressure region mapping between in-shoe and barefoot data.

In-shoe and barefoot pressures were recorded during walking for 25 feet from 16 individuals with high forefoot pressures. High-pressure regions were masked, grouped, and labeled by anatomical region. In-shoe and barefoot pressure maps were averaged over trials and overlaid, and peak pressure and contact area were calculated for all high-pressure regions. For matched regions, centroid locations were compared.

In-shoe high-pressure regions (>200 kPa) were typically observed in similar locations to barefoot walking, while barefoot maps often identified additional regions (>450 kPa) in the toes and lateral forefoot. For in-shoe data, these regions had lower peak pressures and larger contact areas. Matched regions’ centroid positions differed by 0.23–0.60 cm.

Key loading areas were consistent between systems, though unmatched regions were common, particularly in barefoot data, reflecting greater sensitivity to localized pressures. Barefoot data may overestimate high-risk regions that do not display elevated in-shoe pressures. Basing offloading insole design on threshold-based barefoot high-pressure regions could reduce specificity and limit pressure reduction for in-shoe high-pressure regions. These findings have potential implications for streamlining clinical workflows during pressure-informed insole design.

Keywords: Plantar pressure mapping, Custom accommodative insoles, Diabetes

1. Introduction

Foot ulceration is a major concern for people with diabetes, often occurring as a late-stage complication due to factors such as peripheral neuropathy, poor circulation, and repeated stress on the plantar surface from elevated plantar pressures during walking (Chatwin et al., 2020; McDermott et al., 2023). Custom accommodative insoles (CAI) are commonly prescribed to reduce peak plantar pressure (PPP) in regions at risk for ulceration, typically in the forefoot and toes (Bus et al., 2024). Incorporating quantitative biomechanical measures, such as dynamic plantar pressure, in the design process reduces PPP more than insoles designed on shape and subjective clinical assessment (Muir et al., 2025, 2022; Telfer et al., 2017), providing significant protection that reduces ulcer incidence (Bus et al., 2013; Ulbrecht et al., 2014).

A variety of technologies measure dynamic plantar pressure during gait, including both in-shoe and barefoot systems, each with distinct characteristics. In-shoe systems, such as pedar insoles (Novel GmbH, Munich, Germany), use sensor-laden inserts to measure pressure between the foot and shoe, typically with lower spatial resolution, reflecting pressure distribution during in-shoe use. Barefoot systems, such as the emed platform (Novel GmbH, Munich, Germany), measure pressure between the foot and a rigid, flat surface, often providing higher spatial resolution and a more detailed view of intrinsic foot function. Owing to their higher spatial resolution, barefoot systems consistently yield higher PPP values than in-shoe systems when assessed at comparable anatomical locations (Akins et al., 2010; Burnfield et al., 2004; Chockalingam et al., 2025; Owings et al., 2009). In addition, shoes act as a mechanical filter, attenuating forces and pressures such that even with identical sensor resolution, in-shoe values would still be reduced (Huang et al., 2020).

Both in-shoe and barefoot dynamic plantar pressure measurements are used in research to design and optimize CAI. As in-shoe pressure capture technology advances, more studies have incorporated in-shoe pressure data alongside foot shape to guide insole design features such as offloading high-pressure regions (HPRs) (Muir et al., 2025; Waaijman et al., 2012). In studies using both pressure capture systems, barefoot pressure data is often used during insole design, and in-shoe pressure data is used post-fabrication to evaluate pressure reduction (Bus et al., 2004; Owings et al., 2008; Telfer et al., 2017; Ulbrecht et al., 2014; Zwaferink et al., 2020). However, potential differences between barefoot and in-shoe walking pressure distributions, when used to define HPRs for insole design, have not been systematically examined.

While dynamic in-shoe pressure data is more commonly used in research and recommended for custom insole design due to its relevance to real-world use (Owings et al., 2009), clinical implementation remains challenging. In-shoe systems require patients to wear instrumented insoles, cabling, and a control unit, resulting in a more technical and time-intensive setup. In-shoe assessments also require multiple walking steps per trial, increasing data collection time and patient burden. Clinicians must be trained to properly fit the system, verify calibration, and manage data acquisition, creating additional workflow demands. (Carranza et al., 2024). Barefoot pressure data is more efficient to collect, requiring minimal setup and reduced walking time compared to in-shoe pressure collection. If shown to reliably identify the same HPRs as in-shoe data, barefoot measurements could streamline workflows and make pressure-informed insole design more feasible in clinical practice.

While in-shoe and barefoot plantar pressure metrics have been compared holistically in diabetic subjects at prior ulceration sites (Owings et al., 2009), healthy subjects during relaxed standing (Amaro et al., 2025), regionally during gait (Akins et al., 2010; Chockalingam et al., 2025), and at different walking speeds (Burnfield et al., 2004), direct pressure map discrepancies and the exact location of HPRs have not been compared, which have clinical influences on the location of offloading regions during CAI design. Therefore, the aims of this study are to 1) visualize pressure distribution discrepancies of in-shoe and barefoot pressure maps collected from the same subject during walking and 2) determine if HPRs observed during in-shoe and barefoot walking are comparable. PPP, contact area, and the location of HPR centroids in the forefoot will be calculated and compared between the two capture methods.

2. Methods

2.1. Participants

Twenty-five individuals with diabetes (24 males, 1 female) participated in the study. Average (± SD) age was 72.6 ± 6.3 years, height 174.8 ± 6.2 cm, and body mass 94 ± 13.5 kg. Sensation loss by W.H.O. grade was: grade 0 (n = 4), grade 1 (n = 12), and grade 2 (n = 9). Common foot deformities included pes planus (n = 13), claw toes (n = 12), hallux valgus or bony prominences (n = 11), localized edema (n = 11), and equinus calcaneus (n = 5). Frequent comorbidities were prior strokes or heart attacks, hypertension, and hip or knee pain. Inclusion criteria required adults with diabetes, no history of foot ulceration, and the ability to walk independently along an 8 m long walkway without assistive devices for repeated trials. Participants also demonstrated elevated in-shoe PPP in at least one foot, defined as forefoot pressures exceeding 200 kPa, which was verified during testing. Of the 50 feet assessed, 40 met this criterion. Written informed consent was obtained before they participated in this Department of Veterans Affairs Puget Sound Institutional Review Board-approved study (IRB# 1588070).

2.2. Data collection

For the in-shoe walking condition, participants wore standardized research footwear (Dr. Comfort, WI, USA), and in-shoe plantar pressures were acquired using calibrated pedar insoles containing 99 sensor cells (sensels) per insole. The research footwear was extra-deep and wide with the factory insoles removed to minimize the influence of insole geometry and cushioning on the measured in-shoe plantar pressures. Each participant completed five successful walking trials along an 8 m long walkway. For the barefoot condition, plantar pressures were obtained using the emed platform system with a resolution of 4 sensels per cm2 across ten walking trials, capturing five steps per foot. Trial success for both in-shoe and barefoot conditions required walking speeds to be maintained within 10% across trials and between conditions, as measured by infrared timing gates, to minimize differences in spatiotemporal gait parameters. Demographic and health-related information were also recorded. We note that the analyses and results presented here are discussed as in-shoe and barefoot, but the findings are specific to the pressure capture technologies (pedar and emed) used in this study.

2.3. Data analysis

From the 40 feet included in this study, 25 feet (from 16 subjects) were selected based on the presence of in-shoe high-pressure sensels (>200 kPa) at the first, second-to-third metatarsal heads, or a combination of these sites. These regions were identified as the most common areas with high in-shoe plantar pressures. PPP and contact area were calculated using an open-source software package, “pressuRe”, built for the R statistical computing environment and designed to process, analyze, and visualize pressure data (Telfer and Li, 2023).

2.3.1. Aim 1: Overall forefoot comparison

In-shoe and barefoot pressure data differ in sensel resolution and data representation. To enable a direct comparison of forefoot pressure distribution, the raw pressure data was preprocessed to transform the two distinct pressure maps into a comparable format.

Trim Foot to Distal 50%

For both plantar pressure collection methods, PPP was calculated per sensel, and pressure maps (Fig. 1A, 1B) were trimmed to only include the top 50% of the foot to focus on the forefoot and toes.

Fig. 1.

Fig. 1.

Open in a new tab

Pressure data processing pipeline for Aim1: Overall Forefoot Comparison. (A) Raw in-shoe and (B) barefoot pressure maps were trimmed to include distal 50% of the foot. (C) In-shoe spatial internal coordinates are scaled to measured size. Grid logic was applied to barefoot pressure maps while aligned to medial edge of foot, (D) grid IDs were assigned to each sensel per trial, and PPP was averaged across matching grid ID. (E) PPP at or above defined thresholds were set to 1, and values below were linearly scaled from 0 to 1 for both pressure maps. (F) Pressure maps were overlaid to share the same midpoint, and the in-shoe sensel’s PPP was subtracted from each barefoot sensel’s PPP that overlapped the most area.

Average Pressure Over Multiple Steps

For in-shoe walking trials, individual steps were segmented, and PPP per sensel was averaged over all steps taken by the subject across all trials. Raw unitless spatial coordinates from the in-shoe system were scaled to the physical dimensions of each insole, measured using digital calipers (Fig. 1C).

For barefoot walking, multiple trials were aligned and averaged into a single pressure map per foot. We created a global grid based on the maximum number of sensel rows and columns across trials. Barefoot pressure maps were aligned to the medial edge of the foot, and a common global grid was established for all trials per foot. Each sensel was assigned a grid ID per trial (Fig. 1D), and PPP was averaged across matching IDs, ensuring values represented the same foot location across trials.

Scale PPP Across In-Shoe and Barefoot

For averaged in-shoe and barefoot maps, PPP at or above the threshold were assigned 1, while values below were linearly scaled between 0 and 1 (Fig. 1E). Thresholds of 200 kPa (in-shoe) and 450 kPa (barefoot) were chosen based on ulceration risk values in literature (Castro-Martins et al., 2024; Owings et al., 2008; Telfer et al., 2017). This enabled comparison of pressure distributions rather than raw values and reduced the influence of very high PPP in a single sensel.

Overlay Pressure Maps

In-shoe and barefoot pressure maps were aligned at a shared midpoint. The normalized pressure of each barefoot sensel was subtracted from the in-shoe sensel with the greatest overlap (Fig. 1F). This was necessary as some barefoot sensels spanned multiple in-shoe sensels. If a barefoot sensel had no overlap, its difference equaled its own normalized pressure. The final output was the normalized pressure difference per barefoot sensel. Forefoot width was calculated for both systems and compared with ANOVA.

Defining Anatomical Regions

For both in-shoe and barefoot data, the forefoot was divided into 5 anatomical regions (Van der Zwaard et al., 2014): hallux, lesser toes, first metatarsal head (MTH1), second-to-third metatarsal head (MTH2-3), and fourth-to-fifth metatarsal head (MTH4-5) (Fig. 2). For each foot, HPRs were identified where sensels exceeded ulceration risk thresholds and grouped by anatomical region. Left and right feet were analyzed independently, with no mirroring; results are presented as recorded.

Fig. 2.

Fig. 2.

Open in a new tab

Standardized division of the pressure data into 5 anatomically based masking regions for (A) in-shoe and (B) barefoot walking.

2.3.2. Aim 2: Matched HPR locator and mask analysis

To quantifiably compare the HPRs between the two systems, individual barefoot walking trials were preprocessed to rotationally align the foot orientation to in-shoe, and closely spaced HPR masks were combined to enable accurate comparisons to the in-shoe data. HPRs PPP, contact area, and centroid location differences were averaged across barefoot trials for a given foot.

Foot Orientation

For accurate HPR mask location comparison, individual barefoot pressure maps were rotated to vertically align the foot in the same orientation as in-shoe due to small discrepancies in rotation angle between barefoot trials. Barefoot sensel coordinates were rotated to vertically align the midpoint of the heel edge and the most distal sensel in the lesser toes anatomical region (Fig. 3A). Both in-shoe and barefoot pressure maps were then trimmed to only include the top 50% of the foot.

Fig. 3.

Fig. 3.

Open in a new tab

Pressure data processing pipeline for Aim 2: Matched HPR Locator and Mask Analysis. (A) Individual barefoot pressure maps were rotated to align the foot to the in-shoe orientation and trimmed to include the distal 50% of the foot. The green line represents the rotation axis. (B) Separated masks separated by one sensel were combined, circled in red. In-shoe and barefoot pressure maps were then overlaid. (C) Matching HPR centroid location was compared between in-shoe and barefoot. Red dots in the two pressure maps represent the HPR centroid locations being compared. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Combining Separated Masks

Due to sensel masking rules in the “pressure” package, sensels within an HPR mask must share an edge. Barefoot HPRs in the same region but separated by one sensel were not automatically grouped, so manual post-processing merged them into a single mask (Fig. 3B). PPP and contact area were calculated for each original HPR, with combined HPRs assigned the highest PPP and summed contact area. After combining barefoot masks, in-shoe and barefoot maps were overlaid at the same midpoint.

Mask Comparison
Matched HPRs

HPRs in the same anatomical region in both in-shoe and barefoot were classified as matched. For matched pairs, HPR centroids’ mediolateral and proximal–distal location differences were calculated to assess positional shifts, and PPP and contact area values were compared (Fig. 3C). Measures were averaged across barefoot trials for each foot, and ANOVA tested for significant differences between systems. Absolute mediolateral and proximal–distal differences were calculated separately, capturing the magnitude of shifts along each axis without indicating whether they occurred forward/backward or side-to-side. Results for matched and unmatched HPRs were summarized by region.

Unmatched HPRs

Unmatched HPRs appeared in only one system, either in-shoe or barefoot, and were categorized accordingly. For each, PPP and contact area were averaged across trials.

3. Results

3.1. Aim 1: Overall forefoot comparison

When comparing the 25 pairs of in-shoe and barefoot pressure maps, common similarities and differences were observed. Normalized pressure difference was most similar in the central forefoot and lesser toes, but had variability in the lateral forefoot, hallux, and around the periphery of the foot, where barefoot walking captured additional HPRs (Table 1). While areas of high in-shoe PPP were captured during barefoot walking, congruency in pressure distribution varied between feet. Two representative examples are shown to illustrate cases of high (Fig. 4A) and low (Fig. 4B) congruency between in-shoe and barefoot pressure maps. Average forefoot width was significantly narrower in the in-shoe system (9.56 ± 0.53 cm) compared to barefoot (10.17 ± 0.39 cm, p < 0.001).

Table 1.

Count of high-pressure regions by system and region.

Anatomical Region Pressure System
In-shoe (n)		 Barefoot (n)
Hallux 0 10
Lesser Toes 0 11
MTH1 11 20
MTH2-3 23 21
MTH4-5 0 23
Open in a new tab

MTH1 = first metatarsal head; MTH2-3 = second-to-third metatarsal head; MTH45 = fourth-to-fifth metatarsal head.

Fig. 4.

Fig. 4.

Open in a new tab

Normalized pressure difference maps for two representative examples: (A) high congruency and (B) low congruency. Blue sensels indicate lower peak plantar pressure (PPP) during barefoot walking compared to in-shoe, red sensels indicate higher barefoot PPP, and white sensels represent similar PPP values between conditions. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.2. Aim 2: Matched HPR locator and mask analysis

Matched HPRs

Significant differences in PPP and contact area were observed between in-shoe and barefoot matched HPRs in both the MTH1 and MTH2-3 regions. PPP for the MTH1 and MTH2-3 HPRs was significantly higher by 367% and 344% (p < 0.001) in barefoot compared to in-shoe, while contact area was significantly higher by 221% (p = 0.014) and 186% (p = 0.003) in in-shoe compared to barefoot. The mediolateral and proximal–distal location difference of matched HPR centroids was 0.23 cm and 0.60 cm respectively in the MTH1 region, and 0.44 cm and 0.60 cm in the MTH2-3 region (Table 2).

Table 2.

Assessment parameters [Mean (SD)] for matched high-pressure regions by system and region.

Variable Anatomical Region
MTH1 (n = 9)
In-shoe		 Barefoot p-value η2 MTH2-3 (n = 19)
In-shoe		 Barefoot p-value η2
Peak Plantar Pressure (kPa) 235.3 (29.0) 863.5 (239.2) < 0.001 0.79 222.7 (19.4) 765.9 (229.3) < 0.001 0.75
Contact Area (cm2) 4.25 (2.38) 1.92 (0.91) = 0.014 0.32 4.27 (2.34) 2.30 (1.39) = 0.003 0.22
Location Difference
Mediolateral (cm) 0.23 (0.14) 0.44 (0.45)
Proximal–Distal (cm) 0.60 (0.50) 0.60 (0.41)
Open in a new tab

SD = standard deviation; MTH1 = first metatarsal head; MTH2-3 = second-to-third metatarsal head.

Unmatched HPRs

Unmatched in-shoe HPRs were recorded at MTH1 (n = 2) and MTH2-3 (n = 4), with PPP of 203.2 ± 3.2 kPa and 220.3 ± 7.6 kPa and contact areas of 1.98 ± 0.01 cm2 and 5.50 ± 2.83 cm2, respectively. Unmatched barefoot HPRs were recorded at the hallux (n = 10), lesser toes (n = 11), MTH1 (n = 11), MTH2-3 (n = 2), and MTH4-5 (n = 23), with PPP ranging from 553.4 ± 106.3 kPa to 832.3 ± 167.0 kPa and contact areas from 0.28 ± 0.06 cm2 to 2.60 ± 2.26 cm2 (Table 3).

Table 3.

Assessment parameters [Mean (SD)] for unmatched high-pressure regions by system and region.

Anatomical Region Variable
Peak Plantar Pressure (kPa)		 Area (cm^2)
In-shoe
MTH1 (n = 2) 203.2 (3.2) 1.98 (0.01)
MTH2-3 (n = 4) 220.3 (7.6) 5.50 (2.83)
Barefoot
Hallux (n = 10) 616.9 (201.8) 0.57 (0.25)
Lesser Toes (n = 11) 553.4 (106.3) 0.28 (0.06)
MTH1 (n = 11) 691.6 (211.2) 1.01 (0.77)
MTH2-3 (n = 2) 825.0 (376.2) 2.60 (2.26)
MTH4-5 (n = 23) 832.3 (167.0) 0.85 (0.30)
Open in a new tab

SD = standard deviation; MTH1 = first metatarsal head; MTH2-3 = second-to-third metatarsal head; MTH45 = fourth-to-fifth metatarsal head.

4. Discussion

4.1. Aim 1: Overall forefoot comparison

Some feet showed highly similar in-shoe and barefoot pressure distributions, particularly within HPRs, with minimal average normalized differences (Fig. 4A). In contrast, others displayed discrepancies, with barefoot capturing the same HPRs as in-shoe but also revealing additional high-pressure regions, seen as clusters of dark red sensels (Fig. 4B). Additional HPRs were more frequent in barefoot walking, reflecting inherent differences in the pressure measurement systems (Table 1). In-shoe HPRs generally corresponded anatomically with barefoot, though location and contact area varied. This pattern aligns with (Akins et al., 2010), who reported similar regional forefoot pressure rankings (e.g., MTH2 > MTH1 > MTH5) across systems but with lower magnitudes and regional differences in-shoe, consistent with our findings. In contrast, (Owings et al., 2009) found no correlation between in-shoe and barefoot PPP rankings at prior ulceration sites.

4.2. Aim 2: Matched HPR locator and mask analysis

For matched HPRs, the in-shoe system recorded significantly lower PPP and higher contact areas than barefoot, consistent with differences in sensel resolution and foot-shoe interaction. Location differences were minimal, typically < 0.6 cm (Table 2), roughly the size of one barefoot sensel, indicating good spatial agreement. Differences were slightly larger in the proximal–distal direction, likely due to the lower in-shoe resolution, anatomical conformity of insoles, foot migration, gait mechanics, and soft tissue compression (Hurst et al., 2017; Morio et al., 2009). These factors may also explain the high number of unmatched HPRs during barefoot walking (Table 3). Unmatched toe HPRs had lower PPP and smaller contact areas than unmatched MTHs and matched regions, reflecting the barefoot systems’ higher resolution. In the MTH1, unmatched HPRs had higher PPP and contact area than the toes but showed 20% lower PPP and 47% smaller area than matched MTH1 HPRs. The most common unmatched HPR was in the MTH4-5 region, present in 23 of 25 feet during barefoot walking but absent in-shoe. PPP of these HPR resembled matched HPRs in the MTH1 and MTH2-3 regions, while contact area was ~ 59% smaller, likely due to the absence of soft tissue compression in barefoot conditions, which most affects the periphery of the foot (Hurst et al., 2017).

4.3. Factors contributing to pressure differences

System-based differences in pressure sensor resolution, foot-surface interaction, contact mechanics, and shoe compressive effects explain variability in pressure map appearance and number of observed HPRs between barefoot and in-shoe walking. These factors also contribute to discrepancies in matched HPR location, contact areas, and the presence of unmatched barefoot HPRs in the toes and MTH4-5, which have clinical applications in CAI design for those with diabetes and increased ulceration risk.

4.3.1. Spatial resolution differences

Differences in HPR localization, PPP, contact area, and centroid location reflect how each system captures plantar pressure. The emed platform, used in barefoot walking, has higher spatial resolution and a rigid surface, producing concentrated pressure readings allowing small, localized HPRs to be detected accurately (Giacomozzi, 2010). In contrast, in-shoe systems have lower sensel resolution and record pressure through the foot-insole-shoe interface, producing broader, less sharply defined HPRs and greater contact area under regions of high PPP (Abdul Razak et al., 2012; Price et al., 2016). The finer resolution of barefoot systems enables accurate placement of design features, such as metatarsal bars and pads, by capturing forefoot pressure contours underrepresented in in-shoe data. (Telfer et al., 2017; Ulbrecht et al., 2014). In-shoe systems lower resolution may obscure smaller, discrete HPRs that barefoot systems detect. Barefoot pressure maps revealed additional HPRs in the hallux, lesser toes, and MTHs, while in-shoe HPRs were confined to broader regions under MTH1 and MTH2-3 (Table 1). (Owings et al., 2009) showed that barefoot systems, owing to higher resolution, exhibit greater variability in PPP at prior ulceration sites compared to in-shoe systems, where sites were ranked by PPP. In our study, we compared only two regional averages (MTH1 and MTH2-3) rather than ranking all potential HPRs, but the same principle applies. The finer resolution of barefoot measurements captures localized pressure peaks that in-shoe systems often miss. Because in-shoe systems use larger sensors, pressures are averaged across broader areas, yielding less localized values. Consequently, focal HPRs that may be clinically relevant can appear diminished in in-shoe data compared to barefoot recordings.

4.3.2. Foot contact mechanisms

Foot contact differs between in-shoe and barefoot walking. The in-shoe environment transfers pressure through the shoe structure, registering loads even where there is little or no direct foot-floor contact in barefoot walking. This explains higher peripheral PPP in in-shoe maps and broader distribution along the medial and lateral forefoot edges. Barefoot PPP was typically lower at the periphery and in regions with minimal contact, such as between the forefoot and toes, due to natural arching and foot lift during gait. In-shoe foot migration may explain differences in HPR location along the proximal–distal axis. Interactions between foot, insole, shoe, and ground can cause the foot to slip within the shoe, contributing to these location shifts (Morio et al., 2009; Zhu et al., 2024).

4.3.3. Shoe compressive effects

The high number of unmatched HPRs in the MTH4-5 region during barefoot walking may be explained by the inherent soft tissue compression around the periphery of the foot caused by fitting the foot into a standardized shoe size. Observed forefoot width was greater by 0.61 cm in barefoot walking compared to the fitted pedar insole, supporting the idea that the forefoot and toes must compress to fit within the toe box of the shoe, mainly impacting the lateral side of the foot. Reductions in PPP in the MTH4-5 region due to shoe compression are supported by (Zwaferink et al., 2020). In their study, insoles were designed to offload regions with barefoot PPP > 450 kPa using the same offloading techniques across all metatarsal heads. Despite similar barefoot pressures, they found that in-shoe PPPs were 40%–68% less in the MTH4 and MTH5 regions compared to the MTH2 and MTH3. Factors beyond offloading design, such as lateral soft tissue compression during in-shoe walking, likely contribute to these pressure reductions in the MTH4-5 region.

4.4. Clinical implications of unmatched HPRs in in-shoe walking

Unmatched in-shoe HPRs are clinically important, reflecting real-world plantar loading (Owings et al., 2009). Although infrequent, 6 of 25 feet exhibited HPRs in the MTH1 or MTH2-3 during in-shoe walking that were absent barefoot. Notably, the four MTH2-3 HPRs came from two subjects with bilateral findings, potentially indicating altered gait mechanics or more pronounced in-shoe foot migration and tissue compression compared to others in the cohort. Offloading strategies for CAI should consider dynamic foot-shoe interactions to remain effective under functional loading conditions.

5. Limitations

This analysis has several limitations. Barefoot maps sometimes included singular sensels at the foot periphery, contributing to minor spatial discrepancies in medial alignment and global grid, influencing averages across matching grid IDs. Spatial overlay of in-shoe and barefoot data used a midpoint method without scaling scans to share x- and y-axis bounds. Alternative alignment strategies may yield different results and warrant further exploration. Furthermore, in-shoe testing used a single footwear model, which may limit generalizability as pressure patterns could differ across other shoe types. Results cannot be extrapolated to all types of footwear. While directional differences (mediolateral and proximal–distal) were analyzed at the subject level, absolute values were used when averaging across feet to preserve overall magnitude, which may underrepresent directional trends. Pressure thresholds referenced in this study are commonly used but are widely acknowledged to be approximate, with actual safe values likely being region dependent and multifactorial at the patient level. These findings are also system-specific and should not be extrapolated to other plantar pressure technologies without supporting evidence (Telfer and Bigham, 2019).

6. Conclusion

Differences in pressure distribution between in-shoe and barefoot PPP have been previously described (Akins et al., 2010; Amaro et al., 2025; Burnfield et al., 2004), but direct HPR location comparisons are limited. This is clinically relevant, as HPR location directly informs insole offloading, and quantifying spatial differences between in-shoe and barefoot pressure systems is necessary to understand how each approach would differentially influence insole design for high-risk regions. Matched HPRs were found in similar anatomical regions, suggesting key loading areas persist across systems. Unmatched HPRs were common, especially in the lateral forefoot during barefoot walking, reflecting greater sensitivity to localized pressures and shoe compressive effects. These regions highlight a limitation of barefoot data to simulate the real-world shod environment. Barefoot data may overestimate high-risk regions that do not display elevated in-shoe PPP. Basing offloading insole design purely on threshold-based barefoot high PPP regions could reduce specificity and limit pressure reduction for in-shoe high PPP regions. Conversely, failing to include in-shoe HPRs may miss offloading high-pressure areas, with greater clinical consequences. When in-shoe pressures are not feasible to collect, barefoot pressures have the potential to serve as an alternative for identifying in-shoe HPRs, but should be interpreted cautiously, as not all barefoot HPRs persist under in-shoe conditions. This has prospective implications for clinical workflows, streamlining pressure-informed insole design compared to in-shoe pressure collection. Future work comparing insoles designed from in-shoe versus barefoot pressure data utilizing novel workflows could clarify their relative effectiveness in targeting HPRs and outcomes such as reducing PPP at high-risk ulceration sites for those with diabetes, as well as other at-risk populations requiring plantar pressure reduction during gait (Lu et al., 2025).

Acknowledgements

We acknowledge Kimberly A. Nickerson, Mathew Sunil Varre, Christina Carranza, Ellen Y. Li, Ellen Clarissimeaux, and Leo Gagnon from the Department of Veterans Affairs for their contributions to data collection.

Footnotes

CRediT authorship contribution statement

Dylan Heino: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft. Scott Telfer: Writing – review & editing, Supervision, Software, Resources, Methodology, Funding acquisition, Data curation, Conceptualization. Avocet Nagle-Christensen: Writing – review & editing, Investigation. William R. Ledoux: Writing – review & editing, Supervision, Resources, Methodology, Funding acquisition, Conceptualization. Brittney C. Muir: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Abdul Razak AH, Zayegh A, Begg RK, Wahab Y, 2012. Foot plantar pressure measurement system: a review. Sensors (switzerland). 10.3390/s120709884. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Akins JS, Keenan KA, Dugan BP, Francis M, Abt JP, Sell TC, Lephart SM, 2010. Comparison of plantar pressure measurements obtained during barefoot and shod conditions. Med. Sci. Sports Exerc [Google Scholar]
  3. Amaro CM, Paulino MF, Valvez S, Roseiro L, Castro MA, Amaro AM, 2025. Comparative Analysis of pressure Platform and Insole Devices for Plantar pressure Assessment. Applied Sciences (switzerland) 15. 10.3390/app15137575. [DOI] [Google Scholar]
  4. Burnfield JM, Few CD, Mohamed OS, Perry J, 2004. The influence of walking speed and footwear on plantar pressures in older adults. Clin. Biomech 19, 78–84. 10.1016/j.clinbiomech.2003.09.007. [DOI] [PubMed] [Google Scholar]
  5. Bus SA, Sacco ICN, Monteiro-Soares M, Raspovic A, Paton J, Rasmussen A, Lavery LA, van Netten JJ, 2024. Guidelines on the prevention of foot ulcers in persons with diabetes (IWGDF 2023 update). Diabetes Metab. Res. Rev 40. 10.1002/dmrr.3651. [DOI] [PubMed] [Google Scholar]
  6. Bus SA, Ulbrecht JS, Cavanagh PR, 2004. Pressure relief and load redistribution by custom-made insoles in diabetic patients with neuropathy and foot deformity. Clin. Biomech 19, 629–638. 10.1016/j.clinbiomech.2004.02.010. [DOI] [PubMed] [Google Scholar]
  7. Bus Si.A., Waaijman R, Arts M, Haart M. De, Busch-Westbroek T, Van Baal J, Nollet F, 2013. Effect of custom-made footwear on foot ulcer recurrence in diabetes: A multicenter randomized controlled trial. Diabetes Care 36, 4109–4116. DOI: 10.2337/dc13-0996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Carranza C, Nickerson K, Gagnon L, Muir BC, 2024. From Design to Clinical translation: Unraveling Orthotist Perspectives on 3D Printed Accommodative Insoles. JPO Journal of Prosthetics and Orthotics. 10.1097/JPO.0000000000000538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Castro-Martins P, Marques A, Coelho L, Vaz M, Costa JT, 2024. Plantar pressure thresholds as a strategy to prevent diabetic foot ulcers: a systematic review. Heliyon 10. 10.1016/j.heliyon.2024.e26161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chatwin KE, Abbott CA, Boulton AJM, Bowling FL, Reeves ND, 2020. The role of foot pressure measurement in the prediction and prevention of diabetic foot ulceration—A comprehensive review. Diabetes Metab. Res. Rev 10.1002/dmrr.3258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chockalingam N, Giacomozzi C, Healy A, Sacco ICN, 2025. Discrepancies between plantar pressure devices: evaluating cross-system reliability for biomechanics, clinical use and predictive modelling. Foot 64. 10.1016/j.foot.2025.102190. [DOI] [PubMed] [Google Scholar]
  12. Giacomozzi C, 2010. Appropriateness of plantar pressure measurement devices: a comparative technical assessment. Gait Posture 32, 141–144. 10.1016/j.gaitpost.2010.03.014. [DOI] [PubMed] [Google Scholar]
  13. Huang M, Yick KL, Ng SP, Yip J, Cheung RTH, 2020. The effect of support surface and footwear condition on postural sway and lower limb muscle action of the older women. PLoS One 15. 10.1371/journal.pone.0234140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hurst B, Branthwaite H, Greenhalgh A, Chockalingam N, 2017. Medical-grade footwear: the impact of fit and comfort. J Foot Ankle Res 10. 10.1186/s13047-016-0184-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Lu Z, Li X, Sun D, Song Y, Fekete G, Kovács A, Gao Z, Zheng J, Xiang L, Gu Y, 2025. Computationally tuned dual-layer lattice pads adapted to gait-induced pressure distribution. npj Advanced Manufacturing 2. DOI: 10.1038/s44334-025-00055-8. [DOI] [Google Scholar]
  16. McDermott K, Fang M, Boulton AJM, Selvin E, Hicks CW, 2023. Etiology, Epidemiology, and Disparities in the Burden of Diabetic Foot Ulcers. Diabetes Care. 10.2337/dci22-0043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Morio C, Lake MJ, Gueguen N, Rao G, Baly L, 2009. The influence of footwear on foot motion during walking and running. J. Biomech 42, 2081–2088. 10.1016/j.jbiomech.2009.06.015. [DOI] [PubMed] [Google Scholar]
  18. Muir BC, Li JS, Hudak YF, Kaufman GE, Cullum S, Aubin PM, 2022. Evaluation of novel plantar pressure-based 3-dimensional printed accommodative insoles - a feasibility study. Clin. Biomech 98. 10.1016/j.clinbiomech.2022.105739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Muir BC, Nickerson KA, Carranza C, Varre MS, Nagle-Christensen A, Heino D, Li EY, Ledoux WR, Telfer S, 2025. Improving Offloading in Diabetic Feet: Custom 3D-Printed Insoles Outperform Standard of Care. medRxiv. DOI: 10.1101/2025.08.07.25333211. [DOI] [Google Scholar]
  20. Owings TM, Apelqvist J, Stenström A, Becker M, Bus SA, Kalpen A, Ulbrecht JS, Cavanagh PR, 2009. Plantar pressures in diabetic patients with foot ulcers which have remained healed. Diabet. Med 26, 1141–1146. 10.1111/j.1464-5491.2009.02835.x. [DOI] [PubMed] [Google Scholar]
  21. Owings TM, Woerner JL, Frampton JD, Cavanagh PR, Botek G, 2008. Custom Therapeutic Insoles Based on both Foot Shape and Plantar Pressure Measurement Provide Enhanced Pressure Relief. 10.2337/dc07. [DOI] [PubMed] [Google Scholar]
  22. Price C, Parker D, Nester C, 2016. Validity and repeatability of three in-shoe pressure measurement systems. Gait Posture 46, 69–74. 10.1016/j.gaitpost.2016.01.026. [DOI] [PubMed] [Google Scholar]
  23. Telfer S, Bigham JJ, 2019. The influence of population characteristics and measurement system on barefoot plantar pressures: a systematic review and meta-regression analysis. Gait Posture. 10.1016/j.gaitpost.2018.10.030. [DOI] [PubMed] [Google Scholar]
  24. Telfer S, Li EY, 2023. pressuRe: an R package for analyzing and visualizing biomechanical pressure distribution data. Sci. Rep 13. 10.1038/s41598-023-44041-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Telfer S, Woodburn J, Collier A, Cavanagh PR, 2017. Virtually optimized insoles for offloading the diabetic foot: a randomized crossover study. J. Biomech 60, 157–161. 10.1016/j.jbiomech.2017.06.028. [DOI] [PubMed] [Google Scholar]
  26. Ulbrecht JS, Hurley T, Mauger DT, Cavanagh PR, 2014. Prevention of recurrent foot ulcers with plantar pressure-based in-shoe orthoses: the CareFUL prevention multicenter randomized controlled trial. Diabetes Care 37, 1982–1989. 10.2337/dc13-2956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Van der Zwaard BC, Vanwanseele B, Holtkamp F, van der Horst HE, Elders PJM, Menz HB, 2014. Variation in the location of the shoe sole flexion point influences plantar loading patterns during gait. J Foot Ankle Res 7. 10.1186/1757-1146-7-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Waaijman R, Arts MLJ, Haspels R, Busch-Westbroek TE, Nollet F, Bus SA, 2012. Pressure-reduction and preservation in custom-made footwear of patients with diabetes and a history of plantar ulceration. Diabet. Med 29, 1542–1549. 10.1111/j.1464-5491.2012.03700.x. [DOI] [PubMed] [Google Scholar]
  29. Zhu C, Song Y, Xu Y, Zhu A, Baker JS, Liu W, Gu Y, 2024. Toe Box Shape of running Shoes Affects In-Shoe Foot Displacement and Deformation: a Randomized Crossover Study. Bioengineering 11. 10.3390/bioengineering11050457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Zwaferink JBJ, Custers W, Paardekooper I, Berendsen HA, Bus SA, 2020. Optimizing footwear for the diabetic foot: Data-driven custom-made footwear concepts and their effect on pressure relief to prevent diabetic foot ulceration. PLoS One 15. 10.1371/journal.pone.0224010. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES