Review Reports - Intersegmental Coordination Patterns During Heel Rise: Effects of Knee Position and Movement Phases

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study is niche and repetitive. Clinical novelty is limited: the "early/middle/late" segmentation and the flexed vs. extended knee comparison show no evident practical impact.
To provide relevance, a verifiable clinical question should be posed (e.g., the ability of a coordination index to detect dysfunction), and diagnostic utility analyses (sensitivity, specificity, ROC) should be added.
The modified vector coding method is insufficiently described: it lacks a definition of the coupling angle, pattern classification rules, discontinuity management, calculation of proximal/distal dominance, and justification for the temporal division. It is recommended that formulas or pseudocode be made available to facilitate reproducibility.
The sample is minimal and lacks diversity (12 healthy young individuals, normal IPF), limiting the generalizability of the findings: it does not include a pathological midfoot or wide age/sex/BMI ranges.
An a priori sample size calculation justifying n=12 is essential.
The statistical analysis is not entirely clear. The analysis of multiple statistical tests in a small sample size (n) increases the risk of type I error and/or pseudoreplication by treating trials as observations.
The use of a linear mixed model (randomized subject, condition × fixed periods) with False Discovery Rate control and the subject as the unit of analysis is suggested.
As the authors acknowledge, the study has multiple fundamental methodological shortcomings: it does not include midfoot kinematics, EMG, or plantar pressures; knee control is variable.
The conclusions infer muscular changes without measuring them.
Figures should clearly describe the patterns and how "proximal vs. distal dominance" is inferred.
It is suggested that heatmaps indicate the effective n per cell.

Comments on the Quality of English Language

Generally fluent, but there are long sentences, repetitions, and technical terms that require additional precision.

Author Response

Reviewer 1

Comment1

The study is niche and repetitive. Clinical novelty is limited: the "early/middle/late" segmentation and the flexed vs. extended knee comparison show no evident practical impact.

>>Response

Thank you very much for your comment regarding the limited novelty and practical impact of our study. We agree that our work does not directly provide diagnostic utility, and in that sense, its novelty may appear restricted. However, the purpose of this study was to address important knowledge gaps in foot and ankle rehabilitation.

Specifically, although heel-rise tasks are widely used for clinical assessment and treatment, few studies have examined (i) changes in intersegmental coordination across the early, middle, and late phases of elevation, and (ii) the influence of knee position (extended vs. flexed) on such phase-dependent coordination. In clinical practice, it is common to observe patients who report symptoms only in the early or late phase of heel-raise, those who struggle particularly with initiation, or those who show compensatory movements predominantly in the late phase. Despite these observations, research evidence focusing on phase-specific characteristics is scarce.

Moreover, heel-rise is performed not only with the knee extended but also with the knee flexed. Yet, it remains unclear how knee position modulates coordination across different phases. As we noted in the Introduction, adjacent joint positions are known to influence foot stiffness. From this perspective, a phase-specific analysis under different knee conditions has clear clinical significance.

Therefore, rather than relying solely on a single outcome such as “maximum heel height,” we believe that our detailed analysis of coordination according to movement phase and knee condition helps fill an important knowledge gap and provides meaningful insights into subtle changes in foot–ankle control relevant to rehabilitation practice. We clarified these points in the Introduction as follows:

Lines 57-63:

“Previous studies have focused on the maximum elevation of the heel during heel rise [15, 16]. Heel raises are commonly employed in clinical practice; however, the changes occurring in intersegmental coordination across the early, middle, and late phases, or in relation to knee angles, have been scarcely investigated. Despite the frequent observation of phase-specific challenges and compensatory mechanisms in clinical settings, there is a lack of studies directly addressing these phenomena. This gap poses a limitation for conducting individualized assessments and rehabilitation.”　

Lines 79-86:

“This study seeks to address the existing knowledge gaps by offering comprehensive insights into foot and ankle control during heel raises, categorized by phase and knee condition. The findings can contribute to the development of more individualized re-habilitation strategies and to enhance the diagnostic utility of heel raise assessments. In this study, we used the Oxford Foot Model to define four segments of the lower limb and foot: the shank, hindfoot, forefoot, and hallux. This segmentation allowed us to evaluate the intersegmental coordination across the ascending and descending phases.”

Comment 2

To provide relevance, a verifiable clinical question should be posed (e.g., the ability of a coordination index to detect dysfunction), and diagnostic utility analyses (sensitivity, specificity, ROC) should be added.

>>Response

Thank you very much for this important suggestion regarding the need to pose a verifiable clinical question and to include diagnostic utility analyses such as sensitivity, specificity, and ROC curves. In the present study, we focused on healthy participants only, and therefore no pathological groups or clinical outcomes were available that would allow direct calculation of these diagnostic indices.

However, in response to the reviewer’s comment, we have revised the manuscript to explicitly acknowledge this limitation and to frame a future clinical research direction. Specifically, we have added the following statement in Section 4.4 Clinical implications:

Lines 627-635:

“The findings of the present study may enhance the understanding and precision of kinematic assessment and provide preliminary insights into the development of training and rehabilitation strategies for ankle and foot disorders. The validity and reliability of these assessments should be further investigated in future studies. Furthermore, our results provide a foundational reference for phase- and knee-angle–specific coordination patterns in healthy individuals. Building on this foundation, future research should examine whether these coordination indices can detect foot and ankle dysfunction and establish their diagnostic utility (e.g., sensitivity, specificity, ROC analysis) in clinical populations.”

Comment3

The modified vector coding method is insufficiently described: it lacks a definition of the coupling angle, pattern classification rules, discontinuity management, calculation of proximal/distal dominance, and justification for the temporal division. It is recommended that formulas or pseudocode be made available to facilitate reproducibility.

>>Response

Thank you for pointing out that our initial description of the modified vector coding method was insufficient. We agree that the definition of the coupling angle, classification rules, discontinuity management, and justification for temporal division were not fully described. In this revision, we have addressed these issues as follows:

Definition of the coupling angle
In Methods 2.3, we now explicitly state that the coupling angle (γi) was calculated from the differences in angular displacement of the proximal (θ_P) and distal (θ_D) segments between successive frames (θ_P(i+1) – θ_P(i), θ_D(i+1) – θ_D(i)).
Classification rules and proximal/distal dominance
We normalized γi to a 0°–360° range and classified coordination patterns into eight categories (in-phase/anti-phase, proximal/distal dominant, each with ± directions). The detailed classification rules are now provided in Supplementary Table S1.
Justification for temporal division
On Line.144, we clarified that the analysis was based on three temporal divisions (0–33%, 34–66%, 67–100%), corresponding to early, middle, and late phases of the heel-raise. We also added appropriate references to support this segmentation.
Algorithm for reproducibility
To ensure transparency and reproducibility, we have included a Python reference implementation as Supplementary Algorithm S1.

We believe these additions provide sufficient methodological detail and facilitate reproducibility of our analyses.

Comment 4

The sample is minimal and lacks diversity (12 healthy young individuals, normal IPF), limiting the generalizability of the findings: it does not include a pathological midfoot or wide age/sex/BMI ranges.

>> Response

Thank you for your comment. The sample size of 12 participants was determined based on a priori sample size calculation (please see our response to Comment 5 for details). We acknowledge, however, that the limited diversity in age, sex, BMI, and the absence of participants with pathological foot or midfoot conditions reduces the generalizability of our findings. This limitation has been clearly stated in the revised manuscript. In future studies, we plan to recruit larger and more heterogeneous cohorts—including individuals with pathological conditions—to validate and extend the present results.

Lines 606-611:

“Finally, our sample consisted of only 12 healthy young adults with normal foot alignment, all of whom were 21 years old, because most volunteers were recruited from the same undergraduate cohort. Although this homogeneity was not intentional, it further limited the generalizability of our findings. Future studies should recruit larger and more diverse samples, including different age groups, sexes, BMI ranges, and individuals with pathological foot or midfoot conditions, to validate and extend the present results.”

Comment 5

An a priori sample size calculation justifying n=12 is essential.

>> Response

Thank you for your comment. In the present study, the required sample size was estimated using the effect size reported in a previous study (Takabayashi et al.2021) that compared foot range of motion between groups. Using G*Power software (version 3.1.9.4, Kiel University) with effect size (ES) = 0.88, α error probability = 0.05, and power = 0.80, the minimum number of participants required was calculated to be 10. Our sample of 12 participants therefore satisfied this criterion.

We acknowledge that no previous studies have adopted exactly the same design as ours—that is, comparing intersegmental foot coordination under different knee conditions—so it was not feasible to conduct an a priori sample size calculation directly based on this outcome. Therefore, we used the effect size from a related study that compared foot range of motion between groups as a proxy. In future studies, we plan to perform a priori power analyses based on effect sizes derived specifically from intersegmental coordination, and to recruit larger and more diverse samples, including participants with pathological conditions, to validate and extend the present findings. We included the rationale for the sample size in the Methods section as follows:

Lines 109-115:

“Regarding the sample size of the present study, we estimated it based on the effect size reported in a previous study [23] that compared intersegmental coordination between the shank and hindfoot in healthy and flatfoot groups. Using the G*Power software package (version 3.1.9.4, Kiel University), we calculated the required sample size for paired samples under the following conditions: effect size (ES) = 0.88, α error probability = 0.05, and power = 0.80. This calculation indicated that a minimum of 10 participants would be required.”

The small sample size and lack of diversity in age, sex, BMI, and inclusion of pathological foot conditions limit the generalizability of the present findings:

Lines 606-611:

Comment 6

The statistical analysis is not entirely clear. The analysis of multiple statistical tests in a small sample size (n) increases the risk of type I error and/or pseudoreplication by treating trials as observations.

>> Response

Thank you for raising this important point regarding the statistical analysis. In our study, we calculated one representative value from trials. Thus, trials were not treated as separate independent data points, thereby avoiding pseudoreplication.

Furthermore, to reduce the risk of type I error associated with multiple comparisons in a small sample size, we applied Bonferroni correction. These procedures have now been explicitly described in the Methods section to improve transparency and reproducibility as follows:

Lines 183-186:

“A comparative analysis of the heel-rise performance, initial joint configurations, total joint displacement, and coordination patterns across conditions was performed at the participant level, treating repeated trials as within-subject repeated measures rather than independent observations.”

Lines 189-191:

“Post-hoc analysis was performed using paired t-tests or the Wilcoxon signed-rank test with Bonferroni’s significance correction to reduce the risk of Type I error due to multiple comparisons.”

Comment 7

The use of a linear mixed model (randomized subject, condition × fixed periods) with False Discovery Rate control and the subject as the unit of analysis is suggested.

>> Response

Thank you for your valuable suggestion regarding the statistical approach. In this study, participants were treated as the unit of analysis, and repeated trials were not considered independent data points. To account for repeated measures, we used paired tests, repeated-measures ANOVA, or Friedman tests, and we applied Bonferroni correction to reduce the risk of type I error associated with multiple comparisons. This approach was chosen to avoid pseudoreplication and to maintain statistical rigor given the limited sample size.

We fully acknowledge that linear mixed models (LMMs) and False Discovery Rate (FDR) control provide powerful alternatives, particularly when dealing with larger sample sizes, more complex hierarchical structures, or a greater number of conditions. Although the relatively small sample size and limited number of conditions in the present study led us to adopt Bonferroni correction within a repeated-measures design, we agree that LMMs and FDR adjustment would be highly appropriate in future studies with larger and more diverse datasets.

In the revised manuscript, we have explicitly stated the statistical unit of analysis and our use of Bonferroni correction to address multiple comparisons, thereby clarifying our methodological choices in response to the reviewer’s comment.

Lines 183-186:

Lines 189-191:

Comment 8

As the authors acknowledge, the study has multiple fundamental methodological shortcomings: it does not include midfoot kinematics, EMG, or plantar pressures; knee control is variable.

>>Response

Thank you for highlighting the methodological limitations of our study. The foot model employed in this research was restricted to three segments (rearfoot, forefoot, and hallux), and therefore midfoot kinematics were not included. Similarly, supplementary physiological and loading measures such as EMG and plantar pressure were not collected. Regarding knee angle, although participants were instructed to maintain a flexed position (30–45°) and practiced sufficiently before data collection, some variability remained, with an average of 35.5° (SD = 17.9), suggesting inter-subject differences.

These points have now been explicitly acknowledged in the Limitations section. As noted there, future studies will aim to incorporate a multi-segment foot model including the midfoot, as well as additional measures such as EMG and plantar pressure, to provide a more comprehensive analysis. We also plan to implement stricter control of knee positioning to reduce variability in future studies.

Lines 594-602

“Third, we characterized foot kinematics using only a motion capture system. We also did not record electromyographic activity or plantar pressure data, which could provide complementary information on foot rigidity and dynamic support mechanisms. Future studies should combine kinematic, EMG, and plantar pressure data to evaluate the heel-rise movement more comprehensively. Fourth, although the participants were instructed to maintain 30–45° knee flexion in the flexed-knee condition, some inter-individual variability in knee control may have occurred. Future studies should ensure stricter knee angle control to validate and extend the present findings.”

Comment 9

The conclusions infer muscular changes without measuring them.

>> Response

Thank you for pointing this out. In the Discussion, we had referred to potential changes in muscle activity under different knee angle conditions based on existing literature and theoretical mechanisms. As correctly noted, however, no direct measurements such as EMG were conducted in this study.

To address this, we have revised the text to avoid definitive statements. References to muscle activity have been rephrased using terms such as “suggest” or “may indicate,” and we have explicitly stated that EMG or other direct measurements were not included in this study but will be necessary in future research. We made following changes in the revised manuscript:

In Section 4.2 Effect of knee joint configurations on intersegmental coordination, wording regarding muscle activity was changed to non-definitive phrasing (“suggest,” “may indicate”).
On Lines 594–599, we added a clarification that EMG was not measured and should be included in future work. This limitation was also reiterated in Section 4.3 Limitations.

Comment 10

Figures should clearly describe the patterns and how "proximal vs. distal dominance" is inferred.

>> Response

Thank you for this helpful comment. In response, we have added a detailed description of the classification procedure used to determine coordination patterns, including how proximal and distal dominance were inferred (coupling angle calculation definition of each classification). These details are now summarized in Supplementary Table S2.

We made following changes in the revised manuscript:

Supplementary Table S2 now provides the definitions of the eight categories, including the rules for determining proximal vs. distal dominance.
To further clarify, the meaning of the +/– signs (movement direction) has been added directly within the figure legends, with “Inversion (+)” and “Eversion (–)” indicated in each relevant figure.

Comment 11

It is suggested that heatmaps indicate the effective n per cell.

>> Response

Thank you for this suggestion. In our analysis, for each participant and knee condition, 101 coupling angles (γi) were calculated from 0–100% of the time-normalized cycle. Ten trials were averaged, resulting in one representative trial per participant. The time axis was divided into three phases, with 34 points each for Early and Late, and 33 points for Middle (101 points in total per participant). These γi values were then classified into eight coordination patterns based on the definitions provided in Supplementary Table S1. With 12 participants, the total number of data points per phase was 408 for Early/Late and 396 for Middle. These values served as the denominator for calculating the proportion (%) of each coordination pattern. Because directly labeling the counts in the heatmaps would be overly complex, we did not include them in the figures. Instead, we have clarified the analytic procedure and denominators in the Methods section.

We made following changes in the revised manuscript:

Added an explanation of data counts and denominators to the end of Section 2.3 Data analysis (Lines170-182).

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This work investigate to characterize the intersegmental coordination patterns during heel raise, and to evaluate the effect of knee joint configurations on the coordination in both extended and flexed knee conditions. The work is interesting, and claimed to be a reliable reference to improve the clinical assessments and rehabilitation exercises for foot and ankle disorders. The method is generally acceptable, and results are well-explained.

The study includes 12 healthy participants, a sample size too small to provide a statistically reasonable result. An explanation is required or references are needed to proof the sample size is reasonable.

The hypothesis “intersegmental coordination would vary, showing in-phase coordination with shank external rotation and hindfoot inversion, antiphase coordination with hindfoot inversion and forefoot eversion, and in-phase coordination with hindfoot and forefoot plantar flexion during the elevation phase, with opposite effects during descent.” has been stated at the early section. The hypothesis should be addressed in the later part of the manuscript. Is the hypothesis fully or partially verified? What does this indicate? Is the result consist with the hypothesis?

The discussion uses the term “dynamic foot support mechanisms” but provides limited detail on this mechanism.

This work proposes suggestions for rehabilitation, but the experimental studies provided are not rigorously verify the effectiveness. I suggest that suggestions for medical rehabilitation should be made with more caution.

References are way out-dated.

This manuscript makes a valuable contribution to evaluate heel rise biomechanics and its possible clinical applications. More participants with more variety are suggested.

Author Response

We sincerely appreciate the valuable comments from the editor and reviewers, which have greatly contributed to improving the quality of our manuscript. The revised version has also been thoroughly proofread by a native English speaker to ensure accuracy and readability. A point-by-point response to each reviewer’s comment is provided below.

Reviewer 2

Comment 1

>> Response

Thank you very much for your positive evaluation of our study. We are pleased that you found our methods and results to be valid and that you recognized the potential value of this work in contributing to clinical assessment and rehabilitation of foot and ankle disorders. We revised the manuscript according to your comments, which has helped us strengthen and improve the manuscript.

Comment 2

>> Response

Section 2.1 Participants: We added a statement describing the rationale for sample size estimation and its consistency with previous work.
Section 4.3 Limitations: We acknowledged the limited sample size and diversity as limitations, and stated that in future studies we will perform a priori power analyses based on effect sizes derived from intersegmental coordination itself, and recruit larger and more diverse cohorts, including participants with pathological conditions, to validate and extend the present findings.

Lines 109-115:

Lines 606-611:

Comment 3

>> Response

Thank you for this important comment regarding our hypothesis. Our results confirmed that coordination patterns differed between the elevation and descent phases, with opposite patterns observed, which was consistent with our initial hypothesis. As noted in the Results and Discussion, we also detected phase-dependent differences in coordination, supporting the clinical relevance of our findings.

However, the hypothesis that in-phase coordination between shank external rotation and hindfoot inversion would decrease under the flexed knee condition was not supported. In both extended and flexed knee conditions, this in-phase pattern predominated in the early phase, but shifted to antiphase coordination (shank internal rotation and hindfoot inversion with distal dominance) in the middle and late phases. Similarly, our hypothesis that forefoot dominance would increase under the flexed knee condition was not confirmed, as no significant differences were found. Nevertheless, in the early elevation phase, the flexed knee condition showed a significantly higher frequency of proximal-dominant hindfoot plantarflexion with forefoot dorsiflexion (a pattern lowering the medial longitudinal arch), highlighting a distinct movement feature compared to the extended knee condition.

We have also noted that whether these differences are attributable to underlying changes in muscle activity cannot be determined within the scope of this study, and further investigation will be needed.

We made following changes in the revised manuscript:

Added corresponding discussion at the end of Section 4.2 Effect of knee joint configurations on intersegmental coordination. (Lines 570-583)

Comment 4

The discussion uses the term “dynamic foot support mechanisms” but provides limited detail on this mechanism.

>> Response

Thank you for pointing this out. We agree that our description of the “dynamic foot support mechanism” was overly brief. This mechanism refers to the dynamic increase in medial longitudinal arch height and foot stiffness driven by the activity of intrinsic foot muscles as well as extrinsic muscles such as the flexor hallucis longus, flexor digitorum longus, and peroneal muscles. Recent studies have reported that this mechanism contributes more substantially to foot support than the classical Windlass mechanism.

To improve clarity for readers, we have revised the Discussion to include a definition of the dynamic support mechanism along with representative muscle activities. We added explanatory text to Lines.562–567 as follows:

“This reduction of foot rigidity may necessitate a greater contribution from active (or dynamic support) mechanisms to stiffen the medial longitudinal arch, which involve the active engagement of intrinsic and extrinsic foot muscles, including the posterior tibial, anterior tibial, flexor digitorum longus, flexor hallucis longus, and peroneus longus. This mechanism contrasts with the passive Windlass mechanism and has been reported to contribute substantially to foot rigidity during functional tasks [9–11].”

Comment 5

>> Response

Thank you for this important comment. In our manuscript, we referred to the potential clinical applications of our findings; however, as you correctly note, this study did not directly test causal effects or verify the effectiveness of rehabilitation interventions.

In response, we have revised the text to ensure that such statements are positioned as “clinical implications” and “future directions.” Specifically, we added sentences as follows:

“The validity and reliability of these assessments should be further investigated in future studies. We also clarified that our results provide a foundational reference for phase- and knee-angle–specific coordination patterns in healthy individuals, and that future research should examine whether these coordination indices can detect foot and ankle dysfunction and establish their diagnostic utility (e.g., sensitivity, specificity, ROC analysis) in clinical populations.” (Lines 627-635)

Comment 6

References are way out-dated.

>> Response

Thank you for your comment. Regarding the references of Elftman (1960), Hicks (1954), and Alexander (1991), we conducted a review of more recent studies and updated the citations accordingly by incorporating the latest findings. We made following changes in the revised manuscript:

＜修正＞

For Elftman (1960) on the midtarsal joint locking mechanism, we added the following statement reflecting more recent evidence:
“Recent research has revealed that the transverse tarsal joint plays a central role in controlling the foot’s flexibility and rigidity, and that this function arises through coordinated multi-joint, multi-plane motion and muscle activity rather than a simple two-axis model.” (Lines 46–49)
For the other two references, we either updated them to more recent works or removed them from the manuscript.

Comment 7

This manuscript makes a valuable contribution to evaluate heel rise biomechanics and its possible clinical applications. More participants with more variety are suggested.

>> Response

Thank you for this valuable suggestion. We have now explicitly acknowledged the limited number and homogeneity of participants as a methodological limitation of this study. To address this, we added a statement in the Limitations section emphasizing that future studies should recruit a larger sample size and include participants with greater diversity in age, sex, and clinical backgrounds to enhance the generalizability of the findings. (Lines 606-611)

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

Review of Biomechanics article “Intersegmental Coordination Patterns During Heel Rise: Effects of Knee Position and Movement Phases”

General Comments:

The purpose of this study was to examine coordination patterns between the lower leg, hindfoot, forefoot, and hallux during stationary heel raises during the ascending and descending phases. The methodology is strong, although information is needed for clarity. For example, in the introduction, the reader gets to the purpose statement without an understanding of what the segments are in the model. The exact marker placement is not provided or shown. The segment/plane determination for coordination assessment are not stated. The definitions of “proximal dominance” and “distal dominance” are not stated. Figures 1-3 are a good start for organizing the material, but still a bit difficult to understand. The discussion is thorough and well considered.

Specific Comments:

Introduction: Provide cursory details on what segments comprise the foot model so that the reader can comprehend your purpose statement as well as your hypotheses.

Methods, Section 2.1, line 85: It is not clear what is meant by “…and 17 healthy young volunteers who were initially enrolled”. Initially enrolled in what? Is this a sub-study of a bigger study?

Methods, section 2.1: Provide an apriori power analysis of your estimated that sample size and so the reader will know if you met your recruitment goals.

Methods, section 2.2 lines 97-99: Provide a reference for the sentence that begins with “The medial longitudinal arch height ratio …..to foot length”

Methods, section 2.2, line 100. Consider providing a picture of the marker set up. Also, what is meant by “spats”. The spats that I am thinking of are shoe covers that were worn in the early 20^th century. It would not make sense to have the participants wear something that covers their foot. A picture would be quite helpful to the read.

Methods, section 2.2 or 2.3: Please provide more details on how each of the foot segments were defined. What delineates between each section? Do toes 2-5 belong to the forefoot or as they not in a section.

Methods, section 2.3: Specifically state in which segments/planes coordination was assessed. I believe you have at least 4 segments and 3 planes, so the analysis can get complicated quickly.

Table 1: Is the stdev for age really zero? (All subjects were age 21?) I believe that a decimal place is missing in the parenthesis.

Tables 2 and 4: Provide units (degrees) somewhere on each table

Table 4: Why was the shank only studied in the horizontal plane when the other segments were studied in all cardinal planes? It seems that sagittal plane may also be important to know.

Results: Throughout the results section, you mention “proximal dominance” and “distal dominance” but these have not been defined. Do you mean if the proximal segment moved more than the distal segment, it is proximal dominance? What about if they were within a degree or two, so they moved about the same amount? Please define these terms and how they were determined.

Figures 1-3. I think you are on the right track with providing an illustration of the coordination patterns, but I don’t quite understand them yet. I really like the color coding! I also like how you were able to put the statistical significance level in each box. However, I don’t know what the + and – represent. Provide details on the figure caption about which segment is first and which is second on the + /- designation. Perhaps also state that all numbers add to 100% in the 8 boxes in each grouping. (If this is correct, which I think it is). On the Y-axis of each box, where you have “extended” and “flexed” written, can you also add “knee extended condition” and “knee flexed condition” to we know what extended and flexed mean?

Author Response

Reviewer 3

General Comments:

>> Response

Thank you very much for your constructive feedback on the clarity of our methodology. We appreciate your positive evaluation of the overall methodological strength and thoroughness of the Discussion. As you pointed out, several aspects—such as the definition of model segments, marker placement, determination of planes for coordination analysis, and the definitions of proximal vs. distal dominance—required further clarification.

We have carefully revised the manuscript to address these issues and improve readability, while also refining the figures and supplementary materials to enhance clarity. The specific revisions and additions are described in detail in our responses to the “Specific comments” section. We believe that the revision based on your constructive feedback strengthened and improved our manuscript.

Specific Comments:

Comment 1

Introduction: Provide cursory details on what segments comprise the foot model so that the reader can comprehend your purpose statement as well as your hypotheses.

>> Response

Thank you for this helpful suggestion. In the revised manuscript, we have now briefly described the segment composition of the Oxford Foot Model (shank, hindfoot, forefoot, and hallux) so that readers can better understand the purpose statement and hypotheses as follows (Lines 79-86):

“This study seeks to address the existing knowledge gaps by offering comprehensive in-sights into foot and ankle control during heel raises, categorized by phase and knee condition. The findings can contribute to the development of more individualized rehabilitation strategies and to enhance the diagnostic utility of heel raise assessments. In this study, we used the Oxford Foot Model to define four segments of the lower limb and foot: the shank, hindfoot, forefoot, and hallux. This segmentation allowed us to evaluate the intersegmental coordination across the ascending and descending phases.”

Comment 2

>> Response

Thank you for pointing this out. We agree that the phrase “initially enrolled” was unclear. We have revised the text to clarify the relationship between the number of screened individuals and the final number of participants included in the study as follows (Lines101-102):

“Participants were recruited from undergraduate students. In total, 17 healthy young volunteers were initially screened for eligibility.”

Comment 3

Methods, section 2.1: Provide an apriori power analysis of your estimated that sample size and so the reader will know if you met your recruitment goals.

>> Response

Thank you for your comment. In the present study, the required sample size was estimated using the effect size reported in a previous study [Takabayashi et al., 2021] that compared foot range of motion between groups. Using G*Power software (version 3.1.9.4, Kiel University) with effect size (ES) = 0.88, α error probability = 0.05, and power = 0.80, the minimum number of participants required was calculated to be 10. Our final sample of 12 participants therefore satisfied this benchmark.

We also note that no previous studies have adopted exactly the same design as ours—that is, comparing intersegmental foot coordination under different knee conditions—so it was not feasible to conduct a priori sample size calculation directly based on this outcome. For this reason, we adopted the effect size from a related study using foot range of motion as a proxy. In line with your suggestion, we have now included this explanation in Methods, Section 2.1 Participants, so that readers can clearly see how the sample size was estimated and confirm that the recruitment target was achieved.

Nevertheless, we acknowledge that the limited sample size and lack of diversity (age, sex, BMI, and pathological conditions) restrict the generalizability of our findings. Therefore, in Section 4.3 Limitations, we have explicitly stated this point and noted that future studies will conduct a priori power analyses based on effect sizes derived from intersegmental coordination itself, and recruit larger and more diverse cohorts, including participants with pathological conditions, to validate and extend the present results.

Lines 109-115:

Lines 606-611:

Comment 4

Methods, section 2.2 lines 97-99: Provide a reference for the sentence that begins with “The medial longitudinal arch height ratio …..to foot length”

>> Response

Thank you for pointing this out. We agree that a reference was missing for this statement. We added a citation Saltzman et al., 1995 at the relevant sentence in Section 2.2 (Line 122).

Comment 5

>> Response

Thank you for this helpful suggestion. To improve clarity, we have added a figure and table illustrating the marker setup (Supplementary Figure S1, Supplementary Table S1). In addition, we have replaced the ambiguous term “spats” with “short knee-length leggings” to avoid confusion.

Comment 6

>> Response

Thank you for this important comment. In this study, the shank, hindfoot, forefoot, and hallux were defined as independent segments, while toes 2–5 were not modeled as a separate segment. To clarify this, we have added a brief description of each segment definition in Methods (Section 2.2) and provided detailed delineations in Supplementary Figure S1 and Supplementary Table S1. These additions should help readers more clearly understand the segment structure and analytic targets used in our study.

We added Supplementary materials (Figure S1 and Table S1), where detailed definitions of each segment are provided. We also added explanatory text in Section 2.2

Data collection (Lines 123-134) as follows:

“Marker placement followed the Oxford Foot Model protocol [26] (see Supplementary Figure S1 for detailed marker locations). In addition, the shank, hindfoot, forefoot, and hallux were defined as independent segments, whereas toes 2–5 were not included as separate segments (see Supplementary Table S1, Figure S1 for details). The participants were barefoot and wore short knee-length leggings. Three-dimensional joint angles were calculated using a Cardan XYZ rotation sequence (plantarflexion/dorsiflexion, ever-sion/inversion, and abduction/adduction), with each distal segment expressed relative to its adjacent proximal segment. Based on previous studies [18, 23], rearfoot abduc-tion/adduction was expressed as shank internal/external rotation to accurately represent the axial shank rotation. From the resulting kinematic data, the following planes and segments were analyzed: shank rotation in the transverse plane, hindfoot in the sagittal and frontal planes, forefoot in the sagittal and frontal planes, and hallux in the sagittal plane.”

Comment 7

Methods, section 2.3: Specifically state in which segments/planes coordination was assessed. I believe you have at least 4 segments and 3 planes, so the analysis can get complicated quickly.

>> Response

Thank you for this observation. In this study, coordination between the shank and hindfoot was evaluated using shank motion in the transverse plane and hindfoot motion in the frontal plane. Coordination between the hindfoot and forefoot was assessed using frontal plane motion and sagittal plane motion, respectively. We have clarified these details in the Methods (Section 2.3) as follows (Lines 154-159):

Comment 8

Table 1: Is the stdev for age really zero? (All subjects were age 21?) I believe that a decimal place is missing in the parenthesis.

>> Response

Thank you for pointing this out. The standard deviation for age in Table 1 is indeed zero, which is not a typographical error. All participants happened to be 21 years old, as the majority were third-year undergraduate students at our institution. While this was not intentional, it resulted in a highly homogeneous sample, which limits the generalizability of our findings. To address this, we have added a note in the Limitations section emphasizing the restricted age range of participants and the need for future studies to include a broader and more diverse populations (Lines 606-611):

Comment 9

Tables 2 and 4: Provide units (degrees) somewhere on each table

>>Response

Thank you for pointing this out. As suggested, we have added the unit (degrees, °) in the footnotes of Tables 2 and 4 to clarify the angular data.

Comment 10

Table 4: Why was the shank only studied in the horizontal plane when the other segments were studied in all cardinal planes? It seems that sagittal plane may also be important to know.

>> Response

Thank you for this thoughtful comment. In this study, we analyzed shank motion only in the transverse plane based on prior studies (Pohl et al., 2007; Takabayashi et al., 2021), which defined shank rotation relative to hindfoot abduction/adduction as axial (transverse plane) rotation. This approach has been reported to accurately capture axial shank rotation. As also noted in the Introduction, shank rotation is closely coupled with hindfoot motion in the frontal plane, which in turn influences foot stiffness; therefore, our analysis focused on the coordination between these planes.

We acknowledge, however, that sagittal plane shank motion is also clinically relevant. Because such motion is strongly influenced by knee joint movement, we excluded it from the present analysis. Future studies may incorporate sagittal and other planes to provide a more comprehensive understanding of multi-planar coordination.

Comment 11

>> Response

Thank you for raising this point. The definitions and criteria for “proximal dominance” and “distal dominance” were based on the modified vector coding technique (MVCT) described by Needham et al. (2020) and Takabayashi et al. (2021) Using this method, the coupling angle (γi) is calculated and classified into eight coordination patterns in 45° increments (see Supplementary Table S2). In the angle–angle diagram, the proximal segment is plotted on the x-axis and the distal segment on the y-axis, and the slope of the line connecting consecutive data points represents the coupling angle. For example, γi values between 0–45° indicate that changes are greater in the proximal segment (proximal dominance), while values between 45–90° indicate greater changes in the distal segment (distal dominance).

Although a coupling angle of exactly 45° theoretically indicates equal movement of both segments, by convention it is classified into one of the two categories. Thus, even when differences between segments are very small (e.g., within a degree or two), the MVCT necessarily assigns each data point to either proximal or distal dominance. We have clarified this definition in the Methods, Section 2.3 Data Analysis, and provided detailed classification rules in Supplementary Table S2.

Lines 154-169

“Specifically, the coordination between the shank and hindfoot was evaluated using shank rotation in the transverse plane and hindfoot motion in the frontal plane. Coordination between the hindfoot and forefoot was evaluated separately in (1) the sagittal plane (hindfoot plantarflexion/dorsiflexion relative to forefoot plantarflexion/dorsiflexion) and (2) the frontal plane (hindfoot inversion/eversion relative to forefoot inversion/eversion). In accordance with the previous research [27], the coupling angle (γi) was determined by calculating the differences at each time point between the proximal (θP) and distal (θD) segment angles (θP(i+1) – θP(i), θD(i+1) – θD(i)). In the previous study [27], undefined or discontinuous cases arising from zero changes in either segment were handled using piecewise conditions. Similarly, in the present study, specific values (90°, −90°, −180°, or NaN) were assigned to ensure continuity of the coupling angle computation. The com-puted γi was normalized to a range of 0°–360°, and coordination patterns were categorized into eight classifications at 45° intervals (in-phase/anti-phase and proximal/distal domi-nant), as detailed in Supplementary Table S2. To ensure reproducibility, a reference implementation in Python is provided in Supplementary Algorithm S1.”

Lines 171-182

“For each participant and knee condition, 101 time-normalized coupling angles (γi) were obtained across the 0–100% movement cycle. Ten trials were averaged into a single re-presentative trial per participant using circular statistics (mean resultant vector approach) appropriate for angular data to ensure that the periodic nature of γi was maintained. The normalized time series was divided into three periods (early: 0–33%, middle: 34–66%, late: 67–100%), resulting in 34 data points for the early and late periods and 33 data points for the middle period per participant. According to the classification criteria detailed in Supplementary Table S2, each γi was assigned to one of eight coordination patterns. With 12 participants involved, this assignment produced a total of 408 data points per phase (396 during the middle phase). The percentage occurrence of each coordination pattern was then calculated from these data points. To simplify the presentation of the data in the heatmaps, percentages are used instead of absolute counts.”

Comment 12

>> Response

Thank you for this helpful feedback. We agree that additional clarification was needed for Figures 1–3. In response, we have revised the figure captions to include the following details in Figures 1-3:

The “+/–” signs indicate the rotation direction of the proximal and distal segments.
Within each condition and phase, the percentages across the eight boxes sum to 100%.
The labels “Extended” and “Flexed” correspond to the knee extended condition and knee flexed condition, respectively.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I appreciate your effort in reviewing the manuscript. However, the following critical aspects require additional attention:

Unproven clinical relevance: The study lacks a pathological group or diagnostic validation; therefore, I suggest the authors choose between recruiting pilot patients (n=8-10) to demonstrate that their indices detect differences or repositioning it as a normative study by removing all speculative statements about "individualization of rehabilitation."
I suggest removing phrases such as "inform clinical assessments" from the abstract and conclusions, replacing them with "provide preliminary normative data" and "enable future diagnostic validation." Add an explicit section describing what validations are missing before clinical use.
Extreme homogeneity: I suggest stating in the abstract that the 12 participants are exclusively 21-year-old students from a single cohort. Since this is not generalizable to "healthy young adults," but rather a convenience sample with extreme homogeneity, the authors should specify which findings cannot be extrapolated to different age ranges, BMIs, or training levels.
Statistical analysis: I suggest clarifying the number of comparisons performed, as well as the statistical significance value used after the Bonferroni adjustment, and including this information in all results tables, since Table 3 appears not to account for the Bonferroni correction.
The standard deviation of 17.9° for knee angle (average: 35.5°) implies that some participants had a knee angle of 17.6° (barely flexed) and others a knee angle of 53.4° (very flexed), which does not represent a controlled experimental condition. I suggest performing a mandatory analysis of covariance to confirm that variability does not affect the results. Otherwise, it would be necessary to re-collect data with better controls or limit to correlational analysis (coordination vs. knee angle).
Statistical power validation: I suggest calculating observed effect sizes in your data and performing post-hoc statistical power analyses, as the study may be underpowered despite a sufficient sample size calculation if the power is less than 70% for main differences.
Clarity in visualizations: Add notes to heatmaps specifying that each percentage represents observations from 12 participants, as the sample size allows for only one participant to be represented by 8.3%.

Author Response

I appreciate your effort in reviewing the manuscript. However, the following critical aspects require additional attention:

Unproven clinical relevance: The study lacks a pathological group or diagnostic validation; therefore, I suggest the authors choose between recruiting pilot patients (n=8-10) to demonstrate that their indices detect differences or repositioning it as a normative study by removing all speculative statements about "individualization of rehabilitation."

>>Response

Thank you for this valuable comment. We have revised the manuscript to remove speculative statements suggesting that our findings could directly contribute to the individualization of rehabilitation. Instead, we have repositioned the study as a normative study, emphasizing that the present results provide preliminary reference data for future diagnostic and clinical validation. The specific revisions reflecting this change are detailed in the following comments.

I suggest removing phrases such as "inform clinical assessments" from the abstract and conclusions, replacing them with "provide preliminary normative data" and "enable future diagnostic validation." Add an explicit section describing what validations are missing before clinical use.

>>Response:

We sincerely appreciate this constructive suggestion. In accordance with the reviewer’s recommendation, we have carefully removed all expressions implying direct clinical applicability, including “inform clinical assessments,” from the Abstract, Introduction, and Conclusion. These phrases have been replaced with more appropriate and conservative wording that emphasizes the normative and preliminary nature of our findings. Specifically, we now state that the results “provide preliminary normative data” and “may serve as a reference for future diagnostic validation.”

Revised section

Abstract

Lines 12-14:

“This study aimed to provide preliminary normative data on intersegmental coordination patterns during heel rises at different knee joint positions and across various phases and periods.”

Lines 26-29:

“These findings suggest that knee position influences intersegmental coordination during heel rises, and the present results provide reference values that can enable future diagnostic validation and comparative studies in pathological populations.”

Introduction

Lines 80-82:

“Rather than proposing clinical applications, the present investigation provides baseline normative data that may serve as a reference for future diagnostic validation and for comparison with pathological populations.”

Clinical implications

Lines 661-668:

“The present study revealed distinct intersegmental coordination patterns during bilateral heel rise across movement phases and knee angle conditions, providing baseline normative data on coordination strategies in healthy adults. Although the present results do not directly imply clinical applicability, they provide a reference framework for future diagnostic validation and for comparisons with individuals exhibiting foot or ankle dysfunctions. Future research should determine whether these coordination indices can discriminate between healthy and pathological movement patterns and assess their diagnostic performance (e.g., sensitivity, specificity, ROC analysis).”

Conclusions

Lines 672-674:

“These findings provide preliminary normative data that may serve as a reference for future studies aiming to validate diagnostic or rehabilitative applications for foot and ankle disorders.”

Extreme homogeneity: I suggest stating in the abstract that the 12 participants are exclusively 21-year-old students from a single cohort. Since this is not generalizable to "healthy young adults," but rather a convenience sample with extreme homogeneity, the authors should specify which findings cannot be extrapolated to different age ranges, BMIs, or training levels.

>>Response:

We appreciate the reviewer’s insightful comment regarding the sample homogeneity and generalizability of our findings. As suggested, we have explicitly stated in the Abstract, Participants, and Limitations section that all participants were 21-year-old university students from a single cohort, and that this homogeneous sample limits the generalization of the results. However, we would like to note that the present task — a bilateral heel rise — is a simple, low-skill motor task that can be safely performed by individuals across a wide range of ages and physical conditions. Therefore, while minor differences in quantitative measures (e.g., Heel rise performance or total joint displacements) could exist across populations, the qualitative coordination characteristics observed (such as phase-dependent transitions and the influence of knee position) are likely to remain consistent across different age, BMI, and training groups.

Revised section

Abstract

Lines 14-15:

“Twelve 21-year-old university students from the same cohort performed heel rises under knee-extended and knee-flexed conditions.”

2.1. Participants

Lines 100-107:

“Seventeen undergraduate students were initially screened for eligibility. All were 21-year-old university students from the same cohort, recruited to minimize inter-individual variability. We excluded individuals with orthopedic conditions affecting daily activities in the past 6 months, neurological conditions impacting experimental tasks, and abnormal foot posture indicated by a Foot Posture Index (FPI) [22] score outside the range of 0–5. Ultimately, 12 participants (height: 166.3±9.2 cm, body weight: 57.7±9.3 kg; six males; FPI score: 4.1±1.2; right-leg dominant: n=11) were included in the analysis (Table 1). All participants were 21 years old.”

4.3 Limitations

Lines :631-650

“Finally, we only included healthy young adults with normal feet, as verified by the FPI; however, all participants were 21-year-old university students from a single cohort, which limits the generalizability of the results. Nevertheless, the bilateral heel-rise task is a simple and low-skill movement that can be safely performed across a wide range of ages and physical conditions. Therefore, while quantitative measures (e.g., heel rise performance or total joint displacement) may vary with age, BMI, or training level, the qualitative coordination characteristics identified in this study are likely consistent across healthy populations. Moreover, because the present study included only healthy participants, the diagnostic applicability of the proposed indices remains unverified. Future research should validate these indices in clinical populations with foot or ankle pathologies to determine their diagnostic and rehabilitative utility…. Collectively, this study should be interpreted as providing preliminary normative data, which can serve as a reference for future diagnostic validation.”

Statistical analysis: I suggest clarifying the number of comparisons performed, as well as the statistical significance value used after the Bonferroni adjustment, and including this information in all results tables, since Table 3 appears not to account for the Bonferroni correction.

>>Response:

We appreciate the reviewer’s insightful comment. To improve statistical transparency and address this concern, we have revised the Statistical Analysis section to explicitly describe the number of comparisons and the adjusted significance threshold used after Bonferroni correction. Specifically, we now state that post-hoc analyses were conducted using paired t-tests or Wilcoxon signed-rank tests with Bonferroni correction, adjusting the significance level for the three pairwise comparisons among movement periods (adjusted α = 0.05/3 = 0.0167). As Table 3 presents only simple between-condition comparisons, Bonferroni correction was not applied to this table. The correction was applied to the three pairwise period comparisons presented in Table 4, as described in the Statistical Analysis section. We have also added a corresponding note in the Table 4 footnote to clarify this.

Revised section

2.3. Data analysis

Lines 191-196:

“For total joint displacement and intersegmental coordination patterns, with-in-condition comparisons across movement periods (early, middle, and late) were analyzed using ANCOVA (for total joint displacement) or repeated-measures ANO-VA/Friedman tests (for coordination patterns). Post-hoc analyses were performed using paired t-tests or Wilcoxon signed-rank tests with Bonferroni correction, adjusting the significance level for the three pairwise comparisons among the periods (adjusted α = 0.05/3 = 0.0167).”

Table4:

“b: Within-condition comparisons across periods (early, middle, late) were analyzed using ANCOVA with knee flexion angle as a covariate. Post-hoc analyses (Bonferroni-corrected, adjusted α = 0.0167) are described in the main text.”

The standard deviation of 17.9° for knee angle (average: 35.5°) implies that some participants had a knee angle of 17.6° (barely flexed) and others a knee angle of 53.4° (very flexed), which does not represent a controlled experimental condition. I suggest performing a mandatory analysis of covariance to confirm that variability does not affect the results. Otherwise, it would be necessary to re-collect data with better controls or limit to correlational analysis (coordination vs. knee angle).

>> Response:

We appreciate the reviewer’s insightful comment regarding the variability in knee flexion angles. In response, we conducted all subsequent comparisons of heel-rise performance and total joint displacement between the knee-extended (KE) and knee-flexed (KF) conditions using analysis of covariance (ANCOVA), with knee flexion angle included as a covariate to control for inter-individual variability in knee position. As a result, all relevant p-values, effect sizes, and confidence intervals in Tables 3 and 4 have been recalculated based on this adjusted model. The implementation of ANCOVA with knee flexion angle as a covariate resulted in several changes in statistical significance compared with the initial analyses.

Specifically, in the original results, significant between-condition differences were observed in the descending phase for forefoot pronation/supination (late period) and hallux flexion/extension (middle period). After adjusting for knee flexion angle variability, these differences were no longer significant. Conversely, a new significant difference emerged in the descending phase for hindfoot plantar flexion/dorsiflexion (late period). In addition, all previously observed differences in heel-rise performance (Table 3) became non-significant after covariate adjustment. These changes suggest that variability in knee flexion angle partially influenced the initial findings, and that the current ANCOVA-based analysis provides a more robust and reliable interpretation of the condition-related effects.

Revised section

2.3. Data analysis

Lines 187-193:

“However, due to the relatively large variability in knee flexion angles, subsequent comparisons of heel-rise performance and total joint displacement between KE and KF conditions were conducted using analysis of covariance (ANCOVA) with knee flexion angle as a covariate.

For total joint displacement and intersegmental coordination patterns, with-in-condition comparisons across movement periods (early, middle, and late) were analyzed using ANCOVA (for total joint displacement)…”

Table3 and 4:

“Comparisons between conditions (knee-extended vs. knee-flexed) were conducted using analysis of covariance (ANCOVA), with knee flexion angle included as a covariate.”

“F values represent the main effect of condition obtained from the ANCOVA model.”

“Partial η² is reported as an effect size, with 95% confidence intervals (CI) presented as lower and upper bounds.”

Statistical power validation: I suggest calculating observed effect sizes in your data and performing post-hoc statistical power analyses, as the study may be underpowered despite a sufficient sample size calculation if the power is less than 70% for main differences.

>> Response:

We sincerely appreciate this constructive suggestion. In response, we performed post-hoc power analyses using G*Power 3.1 based on the observed effect sizes (partial η²) obtained from the ANCOVA results. The achieved statistical power for the largest observed effect — hindfoot sagittal displacement during the descending phase (partial η² = 0.42) — was 0.745, indicating fair sensitivity for detecting large effects. However, several other comparisons likely exhibited lower power (below 0.70), suggesting that small or subtle effects might not have been detected in the present study. We have accordingly noted this limitation in the revised manuscript (Section 4.3, Limitations), emphasizing the need for future research with larger and more heterogeneous samples to increase statistical power and confirm the robustness of these findings.

Revised section

4.3 Limitations

Lines 641-648:

“Moreover, a post-hoc power analysis was conducted using G*Power 3.1 based on the observed effect sizes (partial η²). The achieved statistical power for the largest effect (hindfoot sagittal displacement during the descending phase, partial η² = 0.42) was 0.745, indicating that the study had fair sensitivity to detect large effects. However, some other comparisons likely exhibited power below 0.70, suggesting that smaller or more subtle effects might not have been detected. Therefore, future studies with larger and more heterogeneous samples are warranted to increase statistical power and confirm the robustness of these findings.”

Clarity in visualizations: Add notes to heatmaps specifying that each percentage represents observations from 12 participants, as the sample size allows for only one participant to be represented by 8.3%.

>> Response:

We appreciate the reviewer’s valuable comment regarding the clarity of the heatmap visualizations. We understand that the percentages in the heatmaps could potentially be misinterpreted as representing the proportion of participants (e.g., one participant = 8.3%). However, in our analysis, each participant contributed multiple data points (33–34 per period) derived from time-normalized coupling angles (γᵢ), each classified into one of eight coordination patterns. Therefore, the percentages shown in the heatmaps represent the relative frequency of each coordination pattern across all observations from 12 participants, rather than the proportion of participants.

Figure 1-3

“Each percentage represents the relative frequency of a coordination pattern calculated across all observations from 12 participants (408 data points per phase, 396 for the middle phase). Because each participant contributed multiple data points per period, percentages reflect the proportion of occurrences among all observations rather than the number of participants.”

Author Response File: Author Response.pdf