Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2026 May 4;16:20512. doi: 10.1038/s41598-026-49858-5

Radiographic Imaging to improve frontal plane weight bearing symmetry and its impact on gait in transtibial bone anchored prosthesis users

Niels Jonkergouw 1,2,✉, Maarten R Prins 1,3, Han Houdijk 2, Manon Witteman 1, Vera G M Kooiman 4, Ruud A Leijendekkers 4
PMCID: PMC13328697  PMID: 42082553

Abstract

Optimal frontal plane alignment of transtibial bone-anchored prostheses (BAPs) is essential to minimize knee joint loading asymmetries. This study examined whether bilateral whole-leg radiographic imaging could improve static alignment and reduce knee joint loading during gait in 23 unilateral transtibial BAP users with press-fit osseointegration implants with an experimental within subject study design. Static weight-bearing line (WBL) alignment and frontal plane knee moments during gait were assessed before and after the intervention. Pain levels were monitored one week before and one week after the intervention. Radiographic imaging significantly improved static WBL symmetry (p < 0.001). Thereby, it can be said with 90% confidence that asymmetries above 4.2 mm can be improved by alignment adjustments. The magnitude of these changes was significantly associated with changes in prosthetic-side peak knee adduction moment (p < 0.001), although no significant group-level improvement in dynamic knee adduction moment asymmetry was observed. Pain levels remained unchanged (p = 0.773), and 19 of 23 participants preferred the radiograph-based alignment over their original setup. These findings support radiographic imaging as an effective method for detecting and correcting static misalignments in transtibial BAP users, while highlighting the need for complementary dynamic gait evaluation.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-49858-5.

Subject terms: Engineering, Health care, Medical research

Introduction

A bone-anchored prosthesis(BAP) is a possible solution for individuals with a lower extremity amputation when activities in daily living are limited due to irreversible socket-related problems, such as skin complications and pain1,2. In lower extremity BAP-users, the prosthetic parts are attached to the human body using a femoral or tibial osseointegration implant3–5. Despite the availability of BAP for over 30 years, there is almost no evidence available concerning optimal frontal-plane alignment of transtibial BAP6. At the same time frontal-plane alignment is reported to be of considerable importance towards prevention and treatment of short- and long-term knee-joint complaints, such as osteoarthritis7,8.

Within prosthesis users, there is the opportunity to influence and optimize the foot´s orientation in respect to the individual’s anatomy. This enables clinicians to manipulate gait parameters, as well as manage joint loading, without surgical interventions. In BAP-users, prosthetic alignment might require different considerations compared to socket-suspended prosthesis-users9. Since the connection between the residual bone and prosthetic socket is not rigid, excessive residual-bone movement within the socket could cause stump-tissue damage10. Therefore, current frontal-plane alignment guidelines focus on optimisation of the socket stability and comfort in conjunction with prosthetic gait11–14. In clinical practice, stabilisation of the stump-socket interface is commonly achieved by reducing the external adduction moments acting on the socket, often through a lateral shift of the prosthetic foot relative to the residual limb comparted to the bench alignment. Such alignment adjustments primary focus is typically on achieving socket comfort and socket-stump stability. Additionally, this could influence knee joint loading by decreasing the external knee adduction moment15, consequently reducing medial knee compartment loading. However, knee joint loading is not always sufficiently considered or systematically evaluated when alignment changes are made to achieve socket stabilisation, which may result in unintended alterations in knee joint loading.

In contrast, an osseointegration implant provides a rigid connection between residual bone and prosthesis’ components. As a result, the focus of prosthetic alignment may be shifted to normalizing joint loading, especially since improper joint loading could result in degenerative changes of the knee joint when walking with a sub-optimal frontal-plane alignment for a prolonged period of time2,8. Consistent with this, patients with osteoarthritis are reported to exhibit significantly higher external peak knee adduction moments16 and interventions targeting these knee adduction moments have been reported to be useful in efforts to delay medial osteoarthritis disease progression17,18. With knee joint varus-valgus malalignments have been linked to degenerative changes of the knee joint which, in certain cases, may even require surgical interventions to be treated7,19–21. Therefore, it would be useful to investigate whether BAP users might benefit from leg alignment interventions aimed at managing knee joint loading, as applied in non-amputated patients with osteoarthritis. This is particularly relevant because BAP users share biomechanical similarities with non-amputated individuals with respect to the rigidity of the connection between the prosthesis and the residual bone, in contrast to socket-suspended prosthetic users.

For people with knee osteoarthritis high-tibial osteotomy and knee arthroplasty22–24 whole-leg radiographic images are used to influence the frontal-plane leg alignment. This is executed by determining the necessary adjustment on a bilateral whole-leg radiographic image with a focus on achieving static leg symmetry. On the radiographic image, a line is drawn from the centre of the hip to the centre of the ankle, called the weight-bearing line(WBL) or mechanical axis7,21,22. Hereafter, the required alignment adjustment in the leg with osteoarthritis is determined while considering the WBL of the contralateral leg21. This approach respects the individual anatomy(by using the contralateral leg as reference) of the patient, respecting the varus or valgus orientation of the knee joint, to select a proper limb alignment after surgery25–27.

In this study we evaluated whether static WBL asymmetry can be reduced in transtibial bone-anchored prosthesis users using whole-leg radiographs and whether such an improvement results in enhanced dynamic symmetry of the knee adduction moments during gait, spatiotemporal gait parameters, and prosthetic comfort. Furthermore, we examined whether prosthetist-executed alignment changes are associated with alterations in peak and overall knee adduction moments during gait, given that excessive moments have been linked to degenerative knee joint changes16,17.

Materials and methods

Participants

This experimental within-subject study recruited unilateral transtibial BAP-users between June 2022 and February 2023, who had undergone implantation with a press-fit osseointegration implant between 2015 and 2021, with a sample size of 27 participants(Appendix I). All eligible participants were approached and included consecutively on a first-come basis, without additional selection, to minimize selection bias. The implants were manufactured by either Orthodynamics GmbHa, AQ Implants GmbHb, or OTN Implants B.V.c. A convenience sampling method was used. Participants eligible for inclusion were over 18 years of age, utilized the BAP for a minimum of one year, could walk without additional assistive devices during measurements as well as during daily prosthetic use. Exclusion criteria included cognitive, communicative, or visual limitations, as well as physical co-morbidities hindering safe locomotion; the unaffected limb was not subject to the exclusion criteria. The study protocol was approved by the Medical Research Ethics Committee Oost-Nederland(NL78861.091.22) and conducted in accordance with the Declaration of Helsinki28. Written informed consent was obtained from all participants prior to their inclusion in the study. The study is registered in the Dutch trial register Overzicht van Medisch-wetenschappelijk Onderzoek in Nederland (OMON) under trial number: NL-OMON52260 (17/05/2022).

Study procedure and parameters

In this study, two conditions were evaluated: the alignment of the daily-living prosthesis, referred to as the initial alignment (IA)-condition, and the alignment based on our intervention, the WBL-condition. Both conditions were evaluated, on the same day, with a bilateral whole-leg radiograph and gait analysis. After the intervention, participants used the WBL-condition for one week. Lower extremity pain was evaluated a week before and after the intervention. Finaly, the participant was asked which alignment was preferred and was provided with that alignment after the experiment.

Radiographic evaluation and alignment intervention

A bilateral whole-leg radiograph was performed with the participant standing upright, wearing their regular prosthesis, including shoes. To standardize stance width, a 10 cm wooden block was manually held between the epicondyles of the knees during acquisition of a comfortable stance. At the same time, a paper, displaying five degrees of external rotation, was positioned in front of the participants’ feet, representing neutral transversal-plane prosthetic-foot alignment.

On the radiographical image, the WBL was drawn from the anatomical rotation centre of the hip to the centre of the ankle for both legs(Fig. 1). The WBL of the non-prosthetic side served as reference, hereafter called target-WBL29. The target-WBL was mirrored onto the prosthetic side, using the hip joint center and distance from the medial epicondyle to the WBL as a reference(Fig. 1). The difference between the WBL and target-WBL was used to determine the necessary prosthetic-foot alignment changes to obtain static leg symmetry21,25,26,30. The asymmetry(in mm) was obtained at two levels: (1) at the level of medial epicondyle to determine the level of WBL asymmetry and (2) below the distal part of the BAP connector to determine the necessary prosthetic alignment change.

Fig. 1.

Fig. 1

Open in a new tab

Determination of distance of weight bearing line to knee joint center and required adjustment. The weight bearing line is obtained from the contralateral leg by drawing a line from the center of the hip to the center of the ankle on the obtained bi-lateral whole leg radiographic image. The perpendicular distance of this weightbearing line to the medial epicondyle is measured as shown in Fig. 1A. Within this particular case: 45.7 mm for the WBL of the prosthesis and 34.5 mm for the non-prothesis side. Hereafter the WBL of the non-prosthetic side(orange line on Fig. 1A) is plotted on the prosthesis, the target WBL. The adjustments to obtain the target WBL alignment was measured at the location of executed adjustment (distal to the bone-anchored prosthesis connector): 15.9 mm. Figure 1B shows the effective change on the prosthetic side at the knee joint after WBL intervention. All WBL values were measured from the medial epicondyles. Note: For clarity purpose, redundant lines were removed with Pixlr Express AI Tool(pixlr.com). No changes were made to the figures itself or the outcome of the data.

If altering the alignment within the existing prosthesis system was not feasible due to built-in height constraints, participants were provided with a Low-Profile Pro-Flexd. A shift of the prosthetic foot could be executed by a shift using a sliding adapter and/or off-set plate as well as with angulations on the prosthetic tube. However, the latter is difficult to reproduce. Therefore, to answer our research questions only participants who underwent a shift(n = 23), were included within this study.

Gait analysis

Gait data was acquired using the M-Gait, an instrumented dual-belt treadmill with embedded force plates and ten high-frequency infrared motion capture camerase. Participants wore 51 reflective markers(39 full-body Plug-In Gait model markers plus 12 calibration markers) which were labelled using Nexus softwaree31. The calibration markers were removed for the dynamic measurements. Marker trajectories were recorded at 100 samples per second and labelled, interpolated and filtered in Vicon Nexus software(version 1.8.4, Vicon MX, Oxford, United Kingdom). The ground reaction forces were recorded at 2000 Hz and filtered with a zero-lag 10 Hz low-pass 4th -order(2 × 2nd order) Butterworth filter.

A warm-up walking trial of approximately five minutes was performed. Then, the comfortable walking speed was determined32. Gait analyses for both the IA-condition and WBL-condition were performed at a comfortable walking speed for two minutes or until a minimum of 25 correct steps per limb were captured33.

Prosthetic comfort and preferred alignment

To understand prosthetic comfort the participants were asked to respond to two questionnaires at the end of each day during two weeks, one week before and one after the intervention. The first gathered the intensity of pain with a Numeric(pain) Rating Scale(NRS) ranging from 0 to 10, the second was to identify the location of the pain(Appendix II)34. At the end of the study, the participant was asked to indicate their preferred alignment.

Data analysis

Static asymmetry

The distance of the WBL to the medial epicondyle on the prosthetic leg was measured on the radiographic images for both conditions. Thereafter, the target-WBL distance, determined on the intact limb in IA-condition, was subtracted from these values. The resulting values represent static asymmetry of the WBLs in millimeters, with 0 representing perfect symmetry (Fig. 1). To determine the WBL asymmetry threshold that could be improved with an alignment intervention, we calculated the minimal correctable asymmetry (MCA) with a 90% confidence as outlined in Appendix III. By estimating the combined errors of our intervention, it was possible to identify who would experience an actual static WBL improvement after the intervention.

Dynamic asymmetry

Marker and force plate data (Plug-in Gait, Vicon) were exported to MATLAB(Release 2021a, The MathWorks Inc., Natick, Massachusetts, United States) and spatiotemporal gait and joint moments data were calculated with the VU 3D model, a rigid-body link segment model35, which defined the frontal plane knee joint center as the mid point between the medial and lateral femoral epicondyles. The mean knee adduction moment (KAM) trajectories were calculated per sample for both conditions (Nm/kg), throughout stance phase(0-100%). Dynamic asymmetry was defined as the absolute difference of the mean KAM trajectories of both legs. On both conditions the root-mean-squared-differences(RMSD) and peak KAM were calculated. With a RMSD of 0 representing perfect dynamic symmetry.

Statistical analysis

Data distribution of discrete values was assessed using a Shapiro-Wilk test, QQ-plots and histograms. Normally distributed data were presented as mean with standard deviation(SD), while non-normally distributed data were reported as median with 25th and 75th percentile. In addition to participant demographics, amputation-specific demographics were used: adjusted body mass index(aBMI)36 and residual limb length37,38. The alpha was set at 0.05, and statistical testing and data visualisation was performed in RStudio39 and MATLAB(Release 2021a)40. Pain before and after the intervention was assessed using a paired-sample t-test.

Static asymmetry

The static asymmetry was tested with a paired-sample t-test on static WBL asymmetry between conditions.

Dynamic asymmetry

The dynamic asymmetry was tested with an one dimensional statistical parametric mapping(SPM-1D) paired sample t-tests to determine whether dynamic KAM asymmetry during stance phase decreased after the alignment intervention by comparing the initial- and WBL-condition within subjects. Additionally, a paired-sample t-test was used to test the RMSD and the spatiotemporal gait parameters.

Relationship of executed alignment to KAM and RMSD change

The relationship of executed alignment (in mm changes) to the change of peak KAM and RMSD-change on the prosthetic side during walking between conditions was tested with a linear regression analysis (r2 and p-value).

Minimal correctible asymmetry

The analyses on dynamic symmetry and relationship of alignment were executed for the complete cohort and the subgroup(with an initial WBL asymmetry above the MCA) separately(Appendix III).

Results

Twenty-three uni-lateral(14 right and 9 left) transtibial BAP-users(17 male and 6 female) were included within this study. The mean ± SD age was 58 ± 11years with a height of 1.80 ± 0.09 m and an adjusted BMI of 29.6 ± 3.4 (Appendix IV).

All participants used an energy-storing and return foot, one of which was combined with a microprocessor(#14, Propio-footd). To achieve the WBL-condition, a low-profile ProFlex footd was required in two participants(#8, 13), while one participant(#5) requested a low-profile ProFlex footd for the study due to personal preference. Fifteen participants used an OPL connectorf, four a GV20f, two a Lucic and two a Nexus g to connect the prosthesis to the osseointegration implant(Appendix IV). Participants walked at a mean preferred walking speed of 4.1 ± 1.0 km/h, with a prosthetic step length of 0.59 ± 0.11 m, a non-prosthetic step length of 0.58 ± 0.09 m and a step width of 0.15 ± 0.04 m, with no significant change of these gait parameters between conditions(Appendix V).

Static asymmetry

Static asymmetry decreased significantly(p < 0.001) from 4.9 mm(IQR: 2.5; 10.3) in IA-condition to 2.0 mm(IQR: 0.8; 2.9) in WBL-condition(Fig. 2). An MCA of 4.2 mm was calculated(Fig. 3). Three participants(#2, 7, 10) did not experienced an improved static asymmetry, two(#7,10) underwent no alignment manipulation as they were already symmetrical in IA-condition and one participant(#2) received an intervention smaller than the MCA, the latter resulted in the same absolute difference of 2 mm between conditions(Fig. 3, Appendix VI).

Fig. 2.

Fig. 2

Open in a new tab

Differences Between WBL and target WBL in both conditions. The difference has been obtained by subtracting the distance(to the medial epicondyle) of target WBL from the WBL. When looking at the top figure it is shown that the executed alignment adjustment could cause an overcorrection, going from a positive to negative value or vice-versa. For that reason, absolute differences to 0 (optimal symmetry) have been calculated, to present the actual difference before and after the intervention. Within participant 2 this absolute difference was exactly the same for both conditions and therefore only one triangle is visual. Light grey represents the initial alignment condition while dark grey is representing weight bearing line condition. The vertical grey lines highlight the participants who did not experience an improved WBL symmetry between conditions.

Fig. 3.

Fig. 3

Open in a new tab

Scatterplot on required alignment change versus achieved alignment change. The y-axis represents the asymmetry value in the initial alignment condition, while the x-axis shows the achieved alignment change in the WBL-condition. Participants within the green area experienced an improvement in static asymmetry between conditions, whereas those in the white area exhibited a decrease in asymmetry in the WBL-condition. The minimal correctable asymmetry was calculated and is visualized by the red dotted lines. The three participants did not experience an improved static symmetry in the WBL-condition. All of these are located within the minimal correctable asymmetry range.

Dynamic asymmetry

The executed alignment change was significantly related with peak KAM change(p < 0.001, r = 0.87, r²=0.75) and RMSD change(p = 0.02, r = 0.48, r²=0.23)(Fig. 4; Table 1). However, SPM showed no significant reduction of the dynamic KAM asymmetry in WBL-condition compared to the IA-condition. Nonetheless, a trend of a reduced KAM in the first half of the stance is visible in the WBL-condition(Fig. 5).

Fig. 4.

Fig. 4

Open in a new tab

Regression Analysis of Absolute Prosthetic Alignment Changes during static WBL alignment and Absolute change in Peak KAM and RMSD during walking. A linear regression analyses examined the relationship between absolute alignment changes and the absolute knee adduction moment amplitude change(Fig. 4 top) and root mean squared difference(RMSD) change(Fig. 4 bottom). The KAM regression analysis reveals that the intercept(-0.007 ± 0.0.009) does not provide a confident estimate of the starting point(p = 0.416). However, the slope(0.007 ± 0.001) indicates a significant relationship(p < 0.001), where each millimeter of alignment change corresponds to a proportional change in knee adduction moment amplitude(Nm/kg). The model explains 75% of the variance(r = 0.87) without adjustment. The bottom figure examines the relationship between absolute alignment change and absolute RMSD change. The intercept(0.016 ± 0.009)(p = 0.083) showed no significant relationship. While the slope(0.002 ± 0.001) indicates a significant relationship(p = 0.02) between alignment changes and RMSD change. This model explains 23% of the variance(r = 0.48). The red dotted line shows the minimal correctible asymmetry. To note, it is of importance to look towards Table 1 to place the RMSD values in context with respect, as the regression analysis shows no significant relationship within the sub-group analysis. Abbreviations: KAM=Knee Adduction Moment, mm=millimeter, Nm/kg=Newton meter / kilogram, RMSD=Root-mean-squared-Difference.

Table 1.

Regression analysis of the absolute alignment change to absolute change of the peak Knee Adduction Moment and Root-Mean-Squared-Difference change.

Complete cohort Cohort with executed alignment larger than MCA
Variable Intercept
Mean(Std. Error)		 P-value Slope
Mean(Std. Error)		 P-value r r-square Intercept
Mean(Std. Error)		 P-value Slope
Mean(Std. Error)		 P-value r r-square
Peak Knee Adduction Moment change(Nm/kg) -0.007(0.009) p = 0.416 0.007(0.001) p < 0.001 0.87 75% 0.008(0.014) p = 0.591 0.007(0.001) P < 0.001 0.90 81%
Root Mean Squared Difference change(Nm/kg) 0.016(0.009) p = 0.083 0.002(0.001) p = 0.020 0.48 23% -0.002(0.024) P = 0.939 -0.002(0.002) P = 0.401 0.26 7%
Open in a new tab

A linear regression analyses was used to examine the relationship between absolute alignment changes and the absolute knee adduction moment amplitude change(Fig. 4A) and root mean squared difference change(Fig. 4B).

Abbreviations: Nm/kg=Newton meter / kilogram.

Fig. 5.

Fig. 5

Open in a new tab

Statistical Parametric Mapping of Knee Moment Asymmetry. The left panel shows the data of the complete cohort, while the right panel is on the subgroup analysis of participants who experienced an intervention larger than the minimal correctible asymmetry. Dynamic asymmetry, was tested with an SPM-1D t-test design, evaluating the absolute difference of the mean KAM trajectories of both legs between conditions. Abbreviations: IA=Initial alignment-condition, Nm/kg=Newton meter/kilogram, SPM=Statistical Parametric Mapping, WBL=Weightbearing line-condition.

The subgroup analysis on participants who underwent a shift larger than MCA of 4.2 mm(n = 13), a stronger significant relationship was found between alignment change and peak KAM change(p < 0.001, r = 0.90, r²=0.81). However, still no significant relationship of executed alignment to RMSD change(p = 0.401) and no significant dynamic symmetry improvement with the SPM analysis was detected(Figs. 4 and 5; Table 1).

Participants pain and preferred prosthetic alignment

One week after intervention no significant change of pain was reported(p = 0.575). Two participants experienced an increase and one a decrease of pain. No change in pain location was experienced for the participants already experiencing pain in IA-condition. Eighty-three%(19/23) preferred to use the WBL-alignment(Table 2).

Table 2.

Pain score before and after intervention and preferred alignment outcome.

Initial Alignment WBL Alignment
Participant Id NRS Pain location NRS Pain location Preferred:
1 0 0 0 0 WBL
2 0 0 0 0 WBL-2
3 0 0 0 0 WBL-4
4 3 5 1 5 WBL
5 0 0 0 0 WBL
6 0 0 0 0 WBL
7 0 0 0 0 WBL
8 0 0 0 0 WBL
9 0 0 0 0 WBL
10 0 0 0 0 WBL
11 0 0 0 0 WBL
12 4 2,3,5 4 2,3,10$$ IA
13 4 1 4 1 IA
14 2 10 # 2 10# WBL + 6
15 0 0 2 10## WBL
16 0 0 0 0 WBL
17 0 0 0 0 WBL
18 0 0 0 0 WBL
19 0 0 0 0 WBL
20 0 0 0 0 WBL
21 0 0 0 0 WBL
22 0 0 0 0 WBL
23 0 0 2 5 WBL + 6
Open in a new tab

All participants monitored their pain score on a NRS-scale one week before and one week after the intervention. This table shows the pain and location of the pain, a zero will be no pain or location. The greyed-out participants experienced a pain score during the study. The last column presents preferred alignment outcome at the end of the study. Sometime small alignment adjustments were required(+ lateral and – a medial shift of the foot in respect to the human body), in relation to technical constrains or participants personal preferences not allowing to obtain the WBL alignment on their own prosthetic foot. Abbreviations: # Lower Back Pain, $$ Pain medially behind patella, ## Femoral head non-prosthetic side. Note: Pain locations have been described in Appendix II.

Discussion

This study demonstrates the feasibility of decreasing static leg WBL-asymmetry in the frontal plane using bilateral whole-leg radiographic image. A significant relationship between absolute alignment adjustment and absolute change in peak KAM during gait was detected, suggesting that prosthetic alignment changes may be useful to mitigate peak KAM, as such peak KAM have been associated to degenerative changes of the knee joint16–18. However, no significant effect on improvements in dynamic asymmetry(mean KAM trajectories) was observed. These results highlights the potential of using radiographic WBL alignment as a tool for improving static limb alignment and mitigate peak knee adduction moments, emphasizing that prosthetic alignment could not solely be based on a static measurement.

In line with the literature, our results show that alignments in current practice demonstrate a wide range of 7 to -24 mm static WBL asymmetry within the IA-condition, where negative values indicate the need for a medial shift of the foot(Appendix VI)11,41. This variability in WBL asymmetry observed in the IA condition reflects current clinical practice, in which bench, static, and dynamic alignment steps are already applied. This is likely due to the reliance on subjective measures such as visual gait assessments and patient-reported comfort, and suggests that the individual’s anatomy in combination with implant placement (orientation may vary within the intramedullary cavity of the tibia) is not considered sufficiently within current clinical alignment approaches. The average improvement in static WBL-asymmetry from 4.9 mm(IQR: 2.5; 10.3) in IA- to 2.0 mm(IQR: 0.8; 2.9) in WBL-condition indicates that radiographic measures could effectively be used to detect excessive static valgus/varus leg asymmetries. As incorrect static knee joint loading is reported to have negative effects, from as early as 18months after detection, potentially such malalignments may lead to degenerative conditions such as knee osteoarthritis within our BAP population7,8,24.

Alignment shifts in our study were correlated with changes in peak KAM, an outcome associated with degenerative knee joint changes17,18. Current clinical practice (this studies initial alignment) did not result in a consistent anatomical starting point and establishing such a guideline appears advisable. Whole-leg radiographs may support this by providing a reproducible reference for static frontal-plane alignment, after which dynamic alignment can be used to fine- tune prosthetic function by subjective gait interpretation. Although such dynamic adjustments will inevitably reintroduce variability, they would do so from a standardized anatomical baseline rather than from a standardized foot positioning as suggested by Blumentritt et al.6. Such an approach would incorporate individual anatomy, including distal femur morphology, implant placement, post-traumatic knee damage, or knee joint disease, and may require clinicians to intentionally deviate from static WBL symmetry based on patient-specific anatomy or pathology.

Within the orthopaedic literature on high-tibial osteotomy and knee arthroplasty, changes on the WBL alignment are used to plan leg alignment and are known to significantly influence peak KAM21,25,26,30. In line with these findings, our results similarly show that prosthetic alignment-induced changes of the WBL were significantly related to changes in peak KAM, supporting the notion that incorporating static measurements could be helpful to mitigate peak dynamic knee-joint loading7,42. At the same time, the SPM analysis shows that the static WBL optimalisation is insufficient to significantly improve dynamic mean KAM asymmetry within transtibial BAP-users over the complete stance phase during gait. Therefore, the proposed WBL radiographic method may serve as a potential approach to monitor and mitigate peak frontal-plane knee joint moments and could be used as a supplementary measure to track alignment changes, particularly as clinical alignment processes may reintroduce static variability over time. These findings further suggest that incorporating dynamic kinetic measurements into the alignment process may be essential for fine-tuning dynamic gait parameters.

The radiographic alignment used in this study appears effective to decrease static WBL asymmetry in 87% of the participants. However, three participants were unresponsive, and notably, two of them showed a change on the radiographic image between IA-condition and WBL-condition despite no alignment adjustment had been made. However, the detected change on participants who underwent no alignment alternation suggests potential measurement errors within our introduced procedure. Therefore, the MCA was assessed, which represents the combined measurement error, radiographic imaging combined with executed alignment, on the prosthetic side. For clinical applicability, leg asymmetry smaller than the MCA of 4.2 mm should be approached with caution. As it remains unclear whether an improvement on static WBL asymmetry, for changes smaller than the MCA, would be a result of the alignment intervention or measurement error.

Participants reported no significant change in prosthetic comfort. However, 83% preferred the WBL-alignment, showing that BAP-users could adapt themselves to the introduced alignment method. At the same time, this is a limitation of this study as the evaluation of prosthetic alignment was conducted over a short-term follow-up period, which is typical in clinical settings but does not account for longer-term outcomes. The inclusion of participants who had been using their BAP for at least one year may also limit the generalizability of the findings, as these well-trained individuals may tolerate a broader range of alignments compared to new BAP-users. Adaptation to the skeletal loading provided by BAP differs considerably from socket-suspended prostheses, and new users may not immediately find WBL symmetry to be comfortable or functional. An additional limitation is that we considered changes in alignment measured at the connector level. Since this does not represent values at the ankle level, it may have contributed to measurement error and should be taken into account in future alignment studies. These considerations should be factored into clinical decision-making.

While frequent radiographic imaging to assess alignment may be impractical and ethically questionable, clinical teams could incorporate WBL alignment evaluations into the routine post-surgical check-ups for osteointegration implant stability. This approach could help ensure static alignment optimalisation while minimizing unnecessary exposure to radiation. However, this study underscores the need for a balanced approach that combines static alignment with dynamic gait analysis, and patient-reported outcomes to optimize dynamic gait.

Conclusion

Our findings indicate that bilateral radiographic images could effectively detect static frontal plane WBL-asymmetry between intact and prosthetic leg and could be used to guide corrective interventions in transtibial BAP alignment. It is advisable to address asymmetries in WBL exceeding 4.2 mm, measured at the knee joint level. A relation between executed alignment change and absolute change in peak KAM was found. However, the intervention did not result in a significant improvement of mean KAM asymmetry during gait. Therefore, fine-tuning alignment during a dynamic evaluation remains essential.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Supplementary Material 1 (419.9KB, docx)

Acknowledgements

The authors would like to acknowledge Alyssa G.M. Groeneveld, MSc and A.T. Sprockel, MSc. for their assistance in the data collection.

Author contributions

N. Jonkergouw : Funding acquisition; Conceptualization; Methodology; Investigation; Formal Analysis; Visualization; Project administration; Resources; Writing - original draft. M.R. Prins : Conceptualization; Software, Methodology; Formal Analysis, Supervision; Writing - review & editing. H. Houdijk: Conceptualization; Methodology; Supervision; Writing - review & editing M. Witteman, Methodology, Formal Analysis, Writing - review & editing V.G.M. Kooiman: Investigation; Software; Data curation; Writing - review & editing. R.A. Leijendekkers: Funding acquisition; Conceptualization; Methodology; Investigation; Project administration; Supervision; Resources; Writing - review & editing.

Funding

This work was supported by the Stichting ZiektenkostenVerzekering Krijgsmacht [grant number 20–0231]. The funding source had no role in the design of this study and did not have any role during its execution, analyses, interpretation of the data, or decision to submit results.

Data availability

The data that support the findings of this study are available from, Ruud Leijendekkers at the Radboudumc but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission from Niels Jonkergouw at the Dutch Military Rehabilitation Centre.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Frölke, J. P. M., Leijendekkers, R. A. & van de Meent, H. Osseointegrierte Prothese für Patienten nach Amputation: Multidisziplinärer Behandlungsansatz in den Niederlanden. Unfallchirurg120, 293–299. 10.1007/s00113-016-0302-1 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Leijendekkers, R. A. et al. Comparison of bone-anchored prostheses and socket prostheses for patients with a lower extremity amputation: a systematic review. Disabil. Rehabil. 39, 1045–1058. 10.1080/09638288.2016.1186752/SUPPL_FILE/IDRE_A_1186752_SM0004.DOCX (2017). [DOI] [PubMed] [Google Scholar]
  • 3.Aschoff, H. H., Clausen, A. & Hoffmeister, T. The endo-exo femur prosthesis–a new concept of bone-guided, prosthetic rehabilitation following above-knee amputation. Z. Orthop. Unfall. 147, 610–615. 10.1055/s-0029-1185893 (2009). [DOI] [PubMed] [Google Scholar]
  • 4.Brånemark, R., Brånemark, P-I., Rydevik, B. & Myers, R. R. Osseointegration in skeletal reconstruction and rehabilitation: A review and the VA San Diego Healthcare System, n.d. [PubMed]
  • 5.Atallah, R. et al. Osseointegrated Transtibial Implants in Patients with Peripheral Vascular Disease. J. Bone Joint Surg.99, 1516–1523. 10.2106/JBJS.16.01295 (2017). [DOI] [PubMed] [Google Scholar]
  • 6.Blumentritt, S. Funktion exoprothetischer Bauteile bei Beinamputierten mit osseointegrierten, perkutan ausgeleiteten Implantaten. Unfallchirurg120, 385–394. 10.1007/s00113-017-0334-1 (2017). [DOI] [PubMed] [Google Scholar]
  • 7.Specogna, A. V. et al. Radiographic Measures of Knee Alignment in Patients with varus Gonarthrosis. Am. J. Sports Med.35, 65–70. 10.1177/0363546506293024 (2007). [DOI] [PubMed] [Google Scholar]
  • 8.Sharma, L. et al. The role of knee alignment in disease progression and functional decline in knee osteoarthritis. J. Am. Med. Assoc.286, 188–195. 10.1001/JAMA.286.2.188 (2001). [DOI] [PubMed] [Google Scholar]
  • 9.Grundei, H. & Timmermann, A. Connecting adapter for coupling exoprostheses to endoshafts: Demands on orthopedic technicians with respect to planning and application. Unfallchirurg120, 378–384. 10.1007/s00113-017-0351-0 (2017). [DOI] [PubMed] [Google Scholar]
  • 10.Jonkergouw, N. et al. Application of ultrasound to monitor in vivo residual bone movement within transtibial prosthetic sockets. Sci. Rep. 2024. 14, 1. 10.1038/s41598-024-60353-7 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Jonkergouw, N., Prins, M. R., Buis, A. W. P. & Van Der Wurff, P. The Effect of Alignment Changes on Unilateral Transtibial Amputee’s Gait: A Systematic Review. PLoS One. 1110.1371/journal.pone.0167466 (2016). [DOI] [PMC free article] [PubMed]
  • 12.Bellmann, M., Blumentritt, S., Pusch, M., Schmalz, T. & Schönemeier, M. Measuring Technology The 3D L. A. S. A. R. – A New Generation of Static Analysis for Optimising Prosthetic and Orthotic Alignment. Orthopädie Technik. 21, 107–13. 10.3109/03093649709164538 (2017).
  • 13.Kobayashi, T., Arabian, A. K., Orendurff, M. S., Rosenbaum-Chou, T. G. & Boone, D. Effect of alignment changes on socket reaction moments while walking in transtibial prostheses with energy storage and return feet. Clin. Biomech. Elsevier Ltd. 29, 47–56. 10.1016/j.clinbiomech.2013.11.005 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kobayashi, T., Orendurff, M. S., Zhang, M. & Boone, D. Individual responses to alignment perturbations in socket reaction moments while walking in transtibial prostheses. Clin. Biomech. Elsevier Ltd. 29, 590–594. 10.1016/j.clinbiomech.2014.04.002 (2014). [DOI] [PubMed] [Google Scholar]
  • 15.Groeneveld, A. M. G. et al. The effect of prosthetic alignment on lower limb kinetics in people with a transtibial bone-anchored prosthesis: An experimental within-subject study. Gait Posture. 117, 274–282. 10.1016/j.gaitpost.2025.01.010 (2025). [DOI] [PubMed] [Google Scholar]
  • 16.Kumar, D., Manal, K. T. & Rudolph, K. S. Knee joint loading during gait in healthy controls and individuals with knee osteoarthritis. Osteoarthr. Cartil.21, 298–305. 10.1016/J.JOCA.2012.11.008 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Chang, A. H. et al. External knee adduction and flexion moments during gait and medial tibiofemoral disease progression in knee osteoarthritis. Osteoarthr. Cartil.23, 1099–1106. 10.1016/j.joca.2015.02.005 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Bennell, K. L. et al. Higher dynamic medial knee load predicts greater cartilage loss over 12 months in medial knee osteoarthritis. Ann. Rheum. Dis.70, 1770–1774. 10.1136/ard.2010.147082 (2011). [DOI] [PubMed] [Google Scholar]
  • 19.Liu, X., Chen, Z., Gao, Y., Zhang, J. & Jin, Z. High Tibial Osteotomy: Review of Techniques and Biomechanics. J. Healthc. Eng.201910.1155/2019/8363128 (2019). [DOI] [PMC free article] [PubMed]
  • 20.Morgenroth, D. C., Medverd, J. R., Seyedali, M. & Czerniecki, J. M. The relationship between knee joint loading rate during walking and degenerative changes on magnetic resonance imaging. Clin. Biomech. Elsevier Ltd. 29, 664–670. 10.1016/J.CLINBIOMECH.2014.04.008 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Park, J. G., Il Bin, S., Kim, J. M. & Lee, B. S. Using the Lower Limb Adduction Angle to Predict Postoperative Knee Joint-Line Obliquity After Open-Wedge High Tibial Osteotomy. Orthop. J. Sports Med.910.1177/23259671211003991 (2021). [DOI] [PMC free article] [PubMed]
  • 22.Lee, S. H., Lee, O. S., Teo, S. H. & Lee, Y. S. Change in gait after high tibial osteotomy: A systematic review and meta-analysis. Gait Posture. 57, 57–68. 10.1016/J.GAITPOST.2017.05.023 (2017). [DOI] [PubMed] [Google Scholar]
  • 23.Ramazanian, T. et al. Distribution and Correlates of Hip-Knee-Ankle Angle in Early Osteoarthritis and Preoperative Total Knee Arthroplasty Patients. J. Arthroplasty. 37, S170–S175. 10.1016/j.arth.2021.12.009 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Whatling, G. M. et al. High tibial osteotomy results in improved frontal plane knee moments, gait patterns and patient-reported outcomes. Knee Surg. Sports Traumatol. Arthrosc.28, 2872–2882. 10.1007/s00167-019-05644-7 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Bellemans, J., Colyn, W., Vandenneucker, H. & Victor, J. The Chitranjan Ranawat award: is neutral mechanical alignment normal for all patients? The concept of constitutional varus. Clin. Orthop. Relat. Res.470, 45–53. 10.1007/S11999-011-1936-5 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Victor, J. M. K. et al. Constitutional varus does not affect joint line orientation in the coronal plane knee. Clin. Orthop. Relat. Res.472, 98–104. 10.1007/s11999-013-2898-6 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Nguyen, H. C. et al. A new protocol for obtaining whole leg radiographs with excellent reproducibility. J. Cartil. Joint Preservation. 210.1016/j.jcjp.2022.100042 (2022).
  • 28.WMA Declaration of Helsinki – Ethical Principles for Medical Research Involving Human Subjects – WMA – The World Medical Association. (n.d.). accessed September 13, (2022). https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/
  • 29.Iseki, Y. et al. Defining the load bearing axis of the lower extremity obtained from anterior-posterior digital radiographs of the whole limb in stance. Osteoarthr. Cartil.17, 586–591. 10.1016/J.JOCA.2008.10.001 (2009). [DOI] [PubMed] [Google Scholar]
  • 30.Miyazaki, T. et al. Dynamic load at baseline can predict radiographic disease progression in medial compartment knee osteoarthritis. Ann. Rheum. Dis.61, 617–622. 10.1136/ARD.61.7.617 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Van Den Bogert, A. J., Geijtenbeek, T., Even-Zohar, O., Steenbrink, F. & Hardin, E. C. A real-time system for biomechanical analysis of human movement and muscle function. Med. Biol. Eng. Comput.51, 1069–1077. 10.1007/s11517-013-1076-z (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Hak, L. et al. Stride frequency and length adjustment in post-stroke individuals: influence on the margins of stability. J. Rehabil Med.47, 126–132. 10.2340/16501977-1903 (2015). [DOI] [PubMed] [Google Scholar]
  • 33.Oliveira, A. S. & Pirscoveanu, C. I. Implications of sample size and acquired number of steps to investigate running biomechanics. Sci. Rep. 2021. 11, 1. 10.1038/s41598-021-82876-z (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Salaffi, F., Stancati, A., Silvestri, C. A., Ciapetti, A. & Grassi, W. Minimal clinically important changes in chronic musculoskeletal pain intensity measured on a numerical rating scale. Eur. J. Pain. 8, 283–291. 10.1016/j.ejpain.2003.09.004 (2004). [DOI] [PubMed] [Google Scholar]
  • 35.Kingma, I., De Looze, M. P., Toussaint, H. M., Klijnsma, H. G. & Bruijnen, T. B. M. Validation of a full body 3-D dynamic linked segment model. Hum. Mov. Sci.15, 833–860. 10.1016/S0167-9457(96)00034-6 (1996). [Google Scholar]
  • 36.Osterkamp, L. Current perspective on assessment of human body proportions of relevance to amputees. J. Am. Diet. Assoc.95, 215–218. 10.1016/S0002-8223(95)00050-X (1995). [DOI] [PubMed] [Google Scholar]
  • 37.International Standards & Organization ISO 8548-2. Prosthetics and orthotics - Limb deficiencies - Part 2: Method of describing lower limb amputation stumps., (1993). https://standards.iteh.ai/catalog/standards/iso/c730c1b6-d5de-4b3a-98cd-b3b9aed57fd7/iso-8548-2-2020
  • 38.Leijendekkers, R. A., van Hinte, G., Nijhuis-van der, M. W., Sanden, J. B. & Staal Gait rehabilitation for a patient with an osseointegrated prosthesis following transfemoral amputation, Physiother. Theory Pract.33, 147–161. 10.1080/09593985.2016.1265620 (2017). [DOI] [PubMed] [Google Scholar]
  • 39.Core Team, R. R: A language and environment for statistical computing. R Foundation for Statistical Computing, (2020). https://www.r-project.org/
  • 40.The MathWorks Inc., The MathWorks Inc. MATLAB version: 2021b, (2023). (2022). https://www.mathworks.com
  • 41.Zahedi, M., Spence, W., Solomonidis, S. & Paul, J. Alignment of lower-limb prostheses. J. Rehabilitation Res.23, 2–19 (1986). [PubMed] [Google Scholar]
  • 42.Jonkergouw, N. et al. The significance of frontal plane static alignment in anticipating dynamic knee moment among transtibial prosthesis users: A cross-sectional study. Gait Posture. 112, 128–133. 10.1016/J.GAITPOST.2024.05.012 (2024). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (419.9KB, docx)

Data Availability Statement

The data that support the findings of this study are available from, Ruud Leijendekkers at the Radboudumc but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission from Niels Jonkergouw at the Dutch Military Rehabilitation Centre.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES