Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Printed Edition

A printed edition of this Special Issue is available at MDPI Books....

Share Help Cite Discuss in SciProfiles

Open AccessFeature PaperArticle

Peer-Review Record

Influence of Swing-Foot Strike Pattern on Balance Control Mechanisms during Gait Initiation over an Obstacle to Be Cleared

Appl. Sci. 2020, 10(1), 244; https://doi.org/10.3390/app10010244

by Romain Artico^1,2,3, Paul Fourcade^1,2, Claudine Teyssèdre^1,2, Teddy Caderby⁴, Arnaud Delafontaine^1,2 and Eric Yiou^1,2,*

Reviewer 1: Anonymous

Reviewer 2: Anonymous

Reviewer 3: Anonymous

Reviewer 4:

Paolo Cavallari

Appl. Sci. 2020, 10(1), 244; https://doi.org/10.3390/app10010244

Submission received: 29 October 2019 / Revised: 17 December 2019 / Accepted: 20 December 2019 / Published: 28 December 2019

(This article belongs to the Special Issue Movement Biomechanics and Motor Control)

Round 1

Reviewer 1 Report

The presented article deals with the question of gait initiation over an obstacle as a functional task for the balance control system. In this task, two swing foot impact patterns were identified. This study investigated the influence of the swing foot strike pattern on the postural organization of the GI over an obstacle to be overcome. An AMTI and a VICON system were used for the investigation.

Attached my remarks

introductory remarks
"More globally, a brief literature review shows that the balance control mechanisms described above have been investigated by independent lines of research (cf. Yiou et al. (2017) for a recent review)."
Rather unspecifically only one source for a far-reaching assertion. One might get the impression the literature research was not sufficiently far-fetched. Authors could disclose their own search criteria to make their own literature more transparent.

To 2.1 Participants. 13 probanten is a very small number for a study. How was the case number calculated?

The data collection and processing for the AMTI was not described. The metrological interaction of all systems remains unclear.

How were the measurement data synchronized with possible different measurement rates? Etc...

2.3 Data analysis
How was the data filtered? 10 and 15hz is a clear usage. 10hz high pass and 15hz low pass? However, this would make no sense as it would limit the information window of the CoP and the measurement results in your information content would lose a lot of information content. Was the beginning of the filtered measurement series excluded from the analysis? (Software filters have charge defects...)

As it seems the CoP was calculated from raw data of the AMTI that would mean that the software of the manufacturer was not used. How was the measurement set up? What was the reason for not using the tracks calculated by the software netforce? Was the "new" measurement system validated?

How was the data of the VICOn system processed?

2.5 Statistics
The significance of an ANOVA for groups of sizes 8 and 5 must be questioned.

In general, it should be noted whether an AMTI equipped with a central torque sensor is the suitable measuring instrument for these investigations. As the horizontal movement over the plate causes longitudinal forces which cannot be detected by the torque sensor. At least the advantages and disadvantages should be discussed against matte ones which measure the planar pressure distribution and a Kistler which measures the forces in all 3 dimensions.

The presentation of the results is dotted and vivid.

Discussion :
The discussion is very detailed but there is no relation or classification of the results in the neurological processes of postural control. Likewise, the influence of the very small group of volunteers remains unmentioned.

Summary:
The summary is very short advised to the authors but they rightlyly state that aging, neurological and musculoskeletal disorders are to be further investigated.

Translated with www.DeepL.com/Translator

Author Response

We wish to thank the reviewer for scrutinizing our manuscript and for the relevant remarks that helped us clarify our thoughts. Please find below the point-to-point reply to each of the comments raised.

Introductory remarks

- "More globally, a brief literature review shows that the balance control mechanisms described above have been investigated by independent lines of research (cf. Yiou et al. (2017) for a recent review)."
Rather unspecifically only one source for a far-reaching assertion. One might get the impression the literature research was not sufficiently far-fetched. Authors could disclose their own search criteria to make their own literature more transparent.

Reply to reviewers. To date, there is no study in the literature that linked foot strike patterns and balance control mechanisms during gait initiation or walking. This fact had been raised in a recent literature review from our group (cf. reference [11]) and justifies the originality of the present research. To substantiate this statement, examples of studies that specifically focused on balance control mechanisms during gait initiation, and examples of studies that specifically focused on foot strike patterns during locomotion are provided in the introduction. We believe that it is not the place in an original article to “disclose search criteria” for the references provided, as it is the case for systematic or scope reviews. We may however remove the reference at this place if the reviewer considers it is not appropriate.

To 2.1 Participants. 13 probanten is a very small number for a study. How was the case number calculated?

Reply. We would like to stress that the present study is not a clinical trial. Consequently, the guidelines stated by the Consort (CONsolidated Standards of Reporting Trials) 2010 for the computation of “case number” or “number need to treat”, are not applicable here. The number of participants involved in the present study is in the classical range of the very large majority of the studies which investigated the posturo-kinetics organization of a voluntary movement in young healthy adults (including virtually all the papers cited in the introduction section). Most of these papers have an impact factor > 2, are classified in the first quartile of Scimago and/or are highly cited in the literature (> 20 citations), which attest to their scientific quality. The use of such relatively low number of participants is due to the fact the biomechanical parameters recorded in these studies are very consistent. Consequently, having thirteen participants performing each ten trials in the two experimental conditions is sufficient to obtain consistent data and to validate specific hypotheses.

The data collection and processing for the AMTI was not described. The metrological interaction of all systems remains unclear.

In the present study, force-plate data were acquired via a 64-channels analogue output board and synchronized with kinematic data using the Vicon acquisition software (Workstation). After data collection, kinematic and force-plate data were exported and processed off-line using a custom-made Matlab program. Specifically, after data filtering (low pass filtering of kinematic and force-plate data), this program converted raw force-plate data (in volts) into physical units (in N for forces and Nm for moments) from the 6x6 calibration matrix provided by the manufacturer (AMTI). Afterwards, the coordinates of the CoP were calculated in accordance with the instructions from the AMTI manufacturer. This is now specified in the revised manuscript (please see page 7, lines 171-174).

How were the measurement data synchronized with possible different measurement rates? Etc...

As above described, force-plate and kinematic signals were simultaneously collected via a 64-channel analog board connected to Vicon motion capture system. In the present study, the sampling frequency was similar for both the force-plate and kinematic data (500 Hz). These signals were integrated and synchronized by the Vicon acquisition software. This is now specified in the revised manuscript (please see page 7, lines 171-174).

It should be noted that he Vicon acquisition software is also capable of synchronizing analog signals acquired with measurement rates different from those of Vicon cameras, provided that the analog sampling rate is an integer multiple of the cameras' frame rate.

In fact, both the kinematic and force-plate data were filtered. Force-plate data were low-pass filtered at 10 Hz and kinematic data were low-pass filtered at 15 Hz. This is now clarified in the revised manuscript (please see page 7, Lines 178-180).

It is noteworthy that these cut-off frequencies have been previously determined in the literature from a residual analysis (Winter, 1990) and have been shown to contain 95 to 99% from the spectral power of these signals (Sinclair et al., 2013), which means that the useful information from these data was preserved after filtering.

Otherwise, in the present study, data acquisition was triggered by the experimenter at least 2 seconds before the "all set" signal. Nevertheless, only the 1500 milliseconds preceding the onset of gait initiation (t0) were included in our analysis, which means that the first filtered data (i.e., at least the first 500 ms) were excluded from the analysis.

Sinclair, J., Taylor, P. J., & Hobbs, S. J. (2013). Digital filtering of three-dimensional lower extremity kinematics: an assessment. Journal of human kinetics, 39, 25–36. doi:10.2478/hukin-2013-0065

Winter DA (1990). Biomechanics and Motor Control of Human Movement, 2nd ed. Wiley, New York.

The coordinates of the CoP have not been computed from the Netforce software, because this software is not designed for the integration and processing of kinematic data acquired from a 3D motion capture system. As above-mentioned, we used the Vicon acquisition software (Nexus) for the integration and synchronized capture of both the kinematic (Vicon) and force-plate data. After data collection, kinematic and force-plate data were exported and processed off-line using a custom-made Matlab program. Specifically, after data filtering (low pass filtering of kinematic and force-plate data), this program converted raw force-plate data (in volts) into physical units (in N for forces and Nm for moments) from the 6x6 calibration matrix provided by the manufacturer (AMTI). Afterwards, the coordinates of the CoP were calculated in accordance with the instructions from the AMTI manufacturer. It is noteworthy that this procedure for processing force-plate data is in accordance with the manufacturer's instructions and consequently does not require validation. This procedure has been used in most of our previous studies (e.g., Caderby et al., 2014; Yiou et al., 2016a; Yiou et al., 2016b). Some methodological details were added to the revised manuscript to clarify this point. Please see page 7, Lines 172-180.

Caderby, T., Yiou, E., Peyrot, N., Begon, M., & Dalleau, G. (2014). Influence of gait speed on the control of mediolateral dynamic stability during gait initiation. Journal of biomechanics, 47(2), 417–423. doi:10.1016/j.jbiomech.2013.11.011

Yiou, E., Fourcade, P., Artico, R., & Caderby, T. (2016). Influence of temporal pressure constraint on the biomechanical organization of gait initiation made with or without an obstacle to clear. Experimental brain research, 234(6), 1363–1375. doi:10.1007/s00221-015-4319-4

Yiou, E., Artico, R., Teyssedre, C. A., Labaune, O., & Fourcade, P. (2016). Anticipatory Postural Control of Stability during Gait Initiation Over Obstacles of Different Height and Distance Made Under Reaction-Time and Self-Initiated Instructions. Frontiers in human neuroscience, 10, 449. doi:10.3389/fnhum.2016.00449

How was the data of the VICOn system processed?

As underlined above, Vicon kinematic data were post-processed (low pass filtered at 15 Hz) and analysed using Matlab. This is now clarified in the revised manuscript. Please see page 7, Lines 172-180.

2.5 Statistics
The significance of an ANOVA for groups of sizes 8 and 5 must be questioned.

Reply. We were not clear with this comment of the reviewer since we did not use “groups of sizes 8 and 5” for the ANOVA, but a single group of thirteen participants. As we stated above, we believe that having thirteen participants performing each ten trials in the two experimental conditions (RFS and FFS) is sufficient to obtain consistent data and to validate specific hypotheses.

It is worth noting that the AMTI force-plate used in our study, like the Kistler force platforms, is a multi-component force platform measuring both the forces and the moments in all 3 dimensions. These two force-plate models consist of four 3-component load cells mounted between the top and the base plate, near the four corners of the platform. Although the nature of these load cells is different (piezoelectric for Kistler vs. strain gauge for AMTI), they allows measuring the 3 force components (Fx, Fy and Fz) and calculating the 3 moment components (Mx, My and Mz), which includes the longitudinal forces caused by the horizontal movement of the body in contact with the platform. Furthermore, the AMTI force platform presents a high sensitivity, low cross-talk and long term stability, which make it ideal for studying human movement within research and clinical contexts. AMTI force platforms are thus classically used for investigating functional tasks, such as gait initiation (Caderby et al., 2014; Caderby et al., 2017; Delafontaine et al., 2017; Yiou et al., 2016a; Yiou et al., 2016b; etc.). Given that, unless special recommendation by the reviewer or the editor, we feel that discussing about the advantages and disadvantages of this force platform is not necessary in the revised manuscript. We will obviously conform to the Editor and Reviewer’s opinion in case a disagreement with the above stated reasons should occur.

Caderby, T., Yiou, E., Peyrot, N., de Viviés, X., Bonazzi, B., & Dalleau, G. (2017). Effects of Changing Body Weight Distribution on Mediolateral Stability Control during Gait Initiation. Frontiers in human neuroscience, 11, 127. doi:10.3389/fnhum.2017.00127

Delafontaine, A., Gagey, O., Colnaghi, S., Do, M. C., & Honeine, J. L. (2017). Rigid Ankle Foot Orthosis Deteriorates Mediolateral Balance Control and Vertical Braking during Gait Initiation. Frontiers in human neuroscience, 11, 214. doi:10.3389/fnhum.2017.00214

The presentation of the results is dotted and vivid.

Reply. The present study is based on a biomechanical analysis and involves young healthy adults, not neurological patients. We therefore not wish to discuss neurological aspects of postural control to avoid potential speculations. As stated in the reply below, our ongoing research investigates the effect of aging and Parkinson’s disease on the postural organization of gait initiation over an obstacle to clear. We believe that it will be more appropriate to discuss the neurological processes of postural control in these papers.

Summary:
The summary is very short advised to the authors but they rightlyly state that aging, neurological and musculoskeletal disorders are to be further investigated.

Reply. The words limit for the abstract is 200 words. In its present form, the abstract reaches this word limit and cannot therefore be extended with an additional sentence related to pathology. As indicated above, our ongoing researches investigate the effect of aging and Parkinson’s disease on the biomechanical organization of gait initiation over an obstacle to clear. This matter will be the scope of a future paper.

Reviewer 2 Report

Revisions:

Methods: The authors need to describe the preferred/natural foot strike pattern over an obstacle for their participants. Line 195-196: It is unclear if l was held constant or was adjusted to the changing distance between the COP and the COM of the person as they take a step. Line 196: Change "corresponded" to "was estimated" Line 199: Insert "merely" between "not fall" Results: insert actual p values throughout, with the exception of (p <0.001) Discussion: The authors have neglected to discuss the continuation of gait after the trailing foot crosses over the obstacle. Seems like the differences (in APAs especially) between foot strike patterns may be more related to the energetic costs of gait propulsion and not just stability related since this study was performed on a young healthy population. Of course, if you are forcing them to perform an unnatural task (see comment 1) then perhaps the change may be stability related. But forcing a person to prioritize stability does not mean that their stability is improved. example: They may have less ML COM/COP motion, but that doesn't mean a person forced to walk a balance beam has improved stability vs. level ground walking. Line 364: Change ", temporal pressure [40] and so on." to ", and temporal pressure [40]." Line 408: I am not sure why "a priori" is here. It seems out of place.

Author Response

Methods:

- The authors need to describe the preferred/natural foot strike pattern over an obstacle for their participants.

Reply: the preferred/natural foot strike pattern was not evaluated in the present study. We however observed in a previous paper on gait initiation made over an obstacle (Yiou et al., 2016) that participants used both the FFS and the RFS patterns in the condition where the obstacle had the same distance/height features as the one used in the present study. No clear preferred/natural foot strike pattern could be detected. As participants of the present study were also healthy young participants, it can therefore be reasonably speculated that they were not much constrained (if any) by the foot strategy imposed in the experiment. Note that for shorter obstacle distance (i.e. distance = 10% of subject’s height), a clear preference for the RFS strategy was systematically observed in participants (Yiou et al. 2016). However, the effect of changing the foot strategy to FFS for such smaller distance was not investigated in the present study. This point is now stressed in the methods (see line 114-117).

Yiou E, Artico R, Teyssedre CA, Labaune O, Fourcade P. (2016) Anticipatory Postural Control of Stability during Gait Initiation Over Obstacles of Different Height and Distance Made Under Reaction-Time and Self-Initiated Instructions. Front Hum Neurosci. 7;10:449.

- Line 195-196: It is unclear if l was held constant or was adjusted to the changing distance between the COP and the COM of the person as they take a step.

Reply: As in our previous studies, l was held constant as it classically done in the literature on locomotion (e.g. Hof et al. 2007, Yiou et al. 2016). It was previously shown that this approximation is valid in human walking (e.g. Hof et al. 2007). This point is added in the text (see line 201).

Hof AL, van Bockel RM, Schoppen T, Postema K. (2007) Control of lateral balance in walking. Experimental findings in normal subjects and above-knee amputees. Gait Posture 25(2):250-8.

- Line 196: Change "corresponded" to "was estimated"

Reply: Done.

- Line 199: Insert "merely" between "not fall"

Reply: Done.

Results:

- insert actual p values throughout, with the exception of (p <0.001)

Reply: Done.

Discussion:

- The authors have neglected to discuss the continuation of gait after the trailing foot crosses over the obstacle. Seems like the differences (in APAs especially) between foot strike patterns may be more related to the energetic costs of gait propulsion and not just stability related since this study was performed on a young healthy population. Of course, if you are forcing them to perform an unnatural task (see comment 1) then perhaps the change may be stability related. But forcing a person to prioritize stability does not mean that their stability is improved. example: They may have less ML COM/COP motion, but that doesn't mean a person forced to walk a balance beam has improved stability vs. level ground walking.

Reply: The present study focused on gait initiation over an obstacle to be cleared. By definition, gait initiation refers to the transient period between quiet standing and the time when the swing (or leading) foot strikes the ground. We conformed to this definition, as we did in our previous articles on gait initiation over an obstacle (e.g. Yiou et al. 2014, 2016), and focused on the biomechanical analyses during this period. Now, following the comment of the reviewer, we went “beyond” this classical time-window since we also report the vertical distance between the obstacle and the rear foot, which corresponds to the safety distance of the rear foot. This variable was not reported in the initial version of the manuscript, but it is now introduced in the methods and in the results section.

In his/her comment, the reviewer further states that because the present study involves young healthy adults (and not a frail population), stability was not an issue and consequently, the reported changes may rather be related to the cost of gait propulsion. We believe that this may not be the case for the following reasons.

1) In the present study, we found that the progression velocity (and therefore the forward COM propulsion) was slightly larger (7%) in the RFS than in the FFS condition. In a previous study (Caderby et al. 2014), we showed that the progression velocity during gait initiation (slow vs. normal vs. rapid) did not influence the mediolateral COM velocity/shift at the time of heel off, foot off and foot contact. It followed that, in the present study, the changes in the mediolateral COM parameters between the RFS and FFS condition can very likely not be ascribed to the slight change in the progression velocity.

2) In walking, the trunk is supported by one leg and the COM “falls” to the contralateral side. One has to develop strategy to cope with this potential instability to ensure safe body progression. This is true whatever the state of the sensory-motor apparatus, be it healthy or not. The development of mediolateral APAs are recognized to be one major strategy to face this instability. There is a growing evidence in the literature showing that young healthy adults adapt the features of these mediolateral APAs to internal (e.g. fatigue, fear of falling), external (e.g. the wear of an orthosis at the stance or swing knee) or environmental constraints applied to the postural system. For example, we showed in a previous study (Yiou et al. 2016) that if these mediolateral APA were not properly scaled with the obstacle features (distance and height), imbalance would occur, as reflected by negative MOS values (Hof et al. 2005). Young healthy adults clearly avoided such imbalance by modulating the amplitude of the mediolateral APAs to reach positive MOS values. In addition, even for tasks such as lateral leg raising (e.g. Gendre et al. 2016) or leg flexion from the erect posture (e.g. Nouillot et al. 200, Hussein et al. 2013) where there is no whole-body progression ‑ and therefore no need for forward COM propulsion‑, the CNS adapts the mediolateral APAs to the constraints applied to the postural system so as to maintain stability.

3) For the reasons evoked in our reply to the first comment of the reviewer, we believe that we did not force participants to perform an unnatural task by asking them to step with a FFS or RFS, since they naturally used these patterns under the condition of obstacle height and distance used in this study.

Finally, we would like to stress that, in the present experiment (as in the experiments of the studies cited above), balance maintenance is indeed probably not really an issue for young healthy participants. We agree with this point raised by the reviewer. But if so, it is precisely because young healthy participants are able to develop optimal postural strategies under conditions with a potential instability, as it is originally shown in the present study (by optimal strategy, we mean strategies requiring the lowest energetic cost).

Caderby T, Yiou E, Peyrot N, Begon M, Dalleau G. (2014) Influence of gait speed on the control of mediolateral dynamic stability during gait initiation. J Biomech. 47(2):417-23.

Gendre M, Yiou E, Gélat T, Honeine JL, Deroche T. (2016) Directional specificity of postural threat on anticipatory postural adjustments during lateral leg raising. Exp Brain Res. 234(3):659-71.

Hof AL, Gazendam MG, Sinke WE. (2005) The condition for dynamic stability. J Biomech. 38(1):1-8.

Hussein T, Yiou E, Larue J. (2013) Age-related differences in motor coordination during simultaneous leg flexion and finger extension: influence of temporal pressure. PLoS One. 8(12):e83064.

Nouillot P, Do MC, Bouisset S. (2000) Are there anticipatory segmental adjustments associated with lower limb flexions when balance is poor in humans? Neurosci Lett. 279(2):77-80.

Line 364: Change ", temporal pressure [40] and so on." to ", and temporal pressure [40]."

Reply: Done.

Line 408: I am not sure why "a priori" is here. It seems out of place.

Reply: “a priori” is removed.

Reviewer 3 Report

This paper presents the results of the experiment carried out to evaluate the effect of swing-foot strike pattern (FFS or RFS pattern) on the postural organisation of GI over an obstacle to be cleared.

The experiment is correctly designed and technically sound. Data and analyses are presented appropriately. Results and conclusions are interesting for the future study of balance control mechanisms in human gait.

Only some mistakes must to be corrected:

There seems to be an error in equation (3) and the corresponding explanation in the text. In the text it is used “Y_COM_FC”, that I suppose it is related to “foot contact”, but in equation and following text it is used “Y_COM_HC”. I think that the subindices are not correct.

In the text, reference numbers should be placed in square brackets [ ], but in the paper some references appear as “Author (year of publication)”. Authors should review all references in the text.

Some references would be completed:

[22] Hatala KG, Dingwall HL, Wunderlich RE, Richmond BG (2013) Variation in Foot Strike Patterns during Running among Habitually Barefoot Populations. PLoS ONE 8(1): e52548. https://doi.org/10.1371/journal.pone.0052548

[45] Hussein, T., Yiou, E., & Larue, J. (2013). Age-related differences in motor coordination during simultaneous leg flexion and finger extension: influence of temporal pressure. PloS one, 8(12), e83064. doi:10.1371/journal.pone.0083064

The “Author Contributions” has not been completed.

Author Response

We wish to thank the reviewer for scrutinizing our manuscript and for his/her positive appreciation that encourages us to continue this research. Please find below the point-to-point reply to each of the comments raised.

Reply. This point is corrected. Thank you.

In the text, reference numbers should be placed in square brackets [ ], but in the paper some references appear as “Author (year of publication)”. Authors should review all references in the text.

Reply. Done. Thank you.

Some references would be completed:

Reply. The DOI for the mentioned references are added.

The “Author Contributions” has not been completed.

Reply. It is completed now.

Reviewer 4 Report

This paper investigates the biomechanics of gait initiation (GI) when bridging a hurdle, a challenging situation for the balance control system. This task may be performed by two alternative strike patterns of the swing-foot: the “rear-foot strike” or the “fore-foot strike”. Participants performed series of GI trials with the instruction to strike the ground either with the rear-foot or with the fore-foot. Results showed that anticipatory postural adjustments in the frontal plane were smaller in amplitude before fore-foot strike than before rear-foot strike, while stability was improved in fore-foot strike. Moreover, it was observed that, at foot contact, the vertical velocity of the centre of mass was higher in fore-foot strike than in rear-foot strike, due to a more effective braking action on the centre of mass in the latter case. However, thanks to the dampening exerted by the fore-foot, the collision forces were smaller in fore-foot strike than in rear-foot strike, as were the slope of these forces and the jerk recorded at the C7 vertebra. In conclusion, authors suggest an interdependent relationship between balance control mechanisms and foot strike pattern to obtain an optimal stability control.

The paper addresses an interesting question. Methods are sound and results seem reliable, however the readability of the text should be improved. In this aim I propose the following comments and amendments.

Major comments

- line 48: At least two main swing-foot strike patterns have been identified in the literature. Are these two patterns age-related in humans?

- lines 82-84: the present study aims to investigate the effect of changing the swing-foot strike pattern (FFS vs. RFS) on ML APAs, on braking of the centre of mass fall, and on related ML stability and force collisions. Even if no significant changes were found in AP direction, nevertheless the paper reports data of both ML and AP directions. So, why the aim declared in the Introduction is limited to ML? By the way, since ML has been used for medio-lateral both in the text and in the figures, please also use AP throughout the text. In this line of thought, it might be worth to decide whether express position and velocity of COM/COP in terms of x-y-z geometrical notation, or into the more straightforward AP-ML-CC (cranial-caudal) anatomical notation, therefore conforming all text and figures accordingly (also for the use of M-P vs. COM-COP). As an example, figure 4 displays ML COM velocity while figure 6 reports z’M for COM velocity along the vertical axis; thus, decide whether using y’M and z’M or using ML COM velocity and CC COM velocity.

- lines 106-107: the participants continued walking with their natural foot pattern after clearing the obstacle. This is a repetition of line 99; is the 5 m track of any importance in this experiment?

- lines 184 and 189: the ML boundary of the base of support is called BOS_y at line 184 and BOS_ymax at line 189. Could you please explain whether these two measures represent the same quantity, or BOS_y should be taken as a relative distance while BOS_ymax as an absolute position limit?

- line 211: immediately after foot contact. Something is wrong (in the text or in the figure) since in fig. 3 the peak and slope of Rz are marked to occur before the foot contact

- line 212: (the “jerk”). This tern is used both to indicate the slope of the derivative of the vertical acceleration of the marker placed at the C7 spine level and to mark the sudden fast change of Rz in figure 3. Please untangle.

- line 223: a one way repeated measures ANOVA with just two levels is more simply known as a paired t-test.

- lines 249, 260 and 264: foot lift duration, safety distance and step length are not significantly different, thus not illustrated, but it would be interesting to know at least their mean values ± SE.

- lines 286-288: it is not clear if the peak value of Rz and its slope are those marked by arrows in figure 3 or not, as those arrows preceded the foot contact.

- lines 316-317: The much larger downward-oriented centre of mass velocity at foot contact in the FFS condition (Δ=74.5%) can be considered a direct consequence of this reduced braking. Figure 6 shows that z’M at FC (i.e. CC COM velocity at FC) is larger in FFS than in RFS, but the plot do not allow to understand if COM is moving cranially (upward) or caudally (downward). By looking at figure 3, one gets even more confused, since COM at FC is moving cranially and with a lower velocity in FFS than in RFS. Moreover, considering the Braking index formula and the fact that at t₀ the z’M should be null, from figure 3 it comes out that the Braking index should be greater than 1, as V_MIN–V_FC > V_MIN. This is clearly in contrast with the results plotted in figure 6, where the average Braking index is lower than 1. Finally, the indication of Braking index in figure 3 is wrong, as it lack the 1/V_MIN factor.

Figures and legends

- Figure 2: why xM is not illustrated?

- Figure 3: When looking at the two bottom panels, it seems that the Rz trace is "contaminated" by the sudden and fast change due to the "Jerk". If this is an artefact due to foot contact, one should conclude that Rz is approximately similar in the two conditions (RFC and FFC) and the "Jerk" is bigger in RFC than FFC,

- Figure 3: in this figure the "Jerk" seems to start just before the heel contact. How is it possible that a collision force occurs before the strike?

- Figure 2 and 3: in these two figures the acronyms TC and HC are introduced to differentiate fore-foot and rear-foot contact times. Authors should be aware that Methods just define foot contact (FC), which is actually used throughout the paper. So, it is likely worth to use FC also in figures 2 and 3.

- Figure 4, 5 and 6: to improve readability, it would be useful to rearrange the plots according to the description order followed in the text. For example, the text first describes bottom and central-right panels of figure 6 (para 3.6), then describes top and central-left panels (para 3.7). So, please decide whether rearrange the figures or the text description.

It would be also useful to change the error bars from +1 SD to +1 SE, as to better illustrate the effects size.

- Figure 4 and 5: please control the results of statistics. The top-left panel of figure 4 and the bottom-right panel of figure 5 report significances, while the text states the contrary.

Minor comments

- lines 19-20: The downward vertical centre of mass velocity was less braked…, you will probably mean The downward velocity of the centre of mass was higher… or The downward fall of the centre of mass was less braked, i.e. its velocity was higher …

- line 33: for parallelism with “postural phase”, also execution phase needs ""

- lines 57-59: It is surprising that the authors of these important lines of research focusing on foot strike pattern effects did not consider that the magnitude of the anticipatory braking of the centre of mass fall might have changed between the RFS and FFS patterns. I would say: It is surprising that the authors of these important lines of research, focusing on foot strike pattern effects, did not consider that the magnitude of the anticipatory braking might have changed between the RFS and FFS patterns.

- line 63: (cf. Yiou et al. (2017) for a recent review). For analogy to the rest of the test should be: (cf. 2, for a recent review)

- line 66: ecological, do you mean "natural"?

- line 78: gait initiating over an obstacle to be cleared … is a very long phrasing. Isn't at this point "gait initiation (GI) over a hurdle" enough?

- line 101: for homogeneity, As in our previous study ¹¹ should be As in our previous study [11]

- line 118: upright standards. Probably upright stands

- line 155: a comma is needed between period and lasting

- line 193 and Eq. 3: all subscripts “HC” should be changed into “FC”

- line 203: you probably mean swing, not stance

- line 235: at the beginning of the swing phase. According to fig. 2 yM reached his peak in the late part of the swing-phase

- lines 256-257: Bars. Do you mean plots?

- line 261 - para 3.4: When the reader reaches this point, he faces the swing phase duration, which is illustrated in the bottom-right panel of figure 6 (not mentioned) and also recalled at line 283, but indicated as execution phase. I suggest, for easiness in reading, to deal with this variable within para 3.7.

Author Response

We wish to greatly thank the reviewer for scrutinizing our manuscript, and for the highly relevant and constructive remarks made on its form and substance. We feel that the manuscript is substantially improved thanks to these remarks and we hope that the amendments we made will satisfy him/her. Please find below the point-to-point reply to each of the comments raised.

Major comments

- line 48: At least two main swing-foot strike patterns have been identified in the literature. Are these two patterns age-related in humans?

Reply: These swing foot patterns have been identified in young healthy adults. It is not known to date whether these patterns are age-dependent. Now, we inform the reviewer that our current research focuses on the effect of aging on the biomechanical organization of gait initiation over an obstacle. Our preliminary results obtained in a group of 10 participants aged between 40-60 years and 10 participants aged between 20-40 years, suggest that the use the FFS pattern increases with aging. This matter will be the subject of a future article in the continuity of the present one.

Reply. We agree. Following the comment of the reviewer, this sentence is reformulated. It is now indicated that “the present study aims to investigate the effect of changing the swing-foot strike pattern (FFS vs. RFS) on the postural organization of gait initiation over an obstacle. “

- By the way, since ML has been used for medio-lateral both in the text and in the figures, please also use AP throughout the text. In this line of thought, it might be worth to decide whether express position and velocity of COM/COP in terms of x-y-z geometrical notation, or into the more straightforward AP-ML-CC (cranial-caudal) anatomical notation, therefore conforming all text and figures accordingly (also for the use of M-P vs. COM-COP). As an example, figure 4 displays ML COM velocity while figure 6 reports z’M for COM velocity along the vertical axis; thus, decide whether using y’M and z’M or using ML COM velocity and CC COM velocity.

Reply : Following the remark of the reviewer, we indeed realized that the terms we used for “mediolateral” and “anteroposterior” were not homogeneous throughout the text and figures. The acronym “ML” and “AP” are now systematically used, after being defined the first time they appear in the text. Please also note that we preferred to employ the term “vertical” rather than “cranial caudal”, because it seems to be less used in the literature to express the components of the braking index. This would also be consistent with our previous studies.

- lines 106-107: the participants continued walking with their natural foot pattern after clearing the obstacle. This is a repetition of line 99; is the 5 m track of any importance in this experiment?

Reply: The repetition is removed. In the present study, we used the same standardized conditions with a 5 m walking track as in our previous studies on gait initiation made with or without an obstacle to clear. With these standardized conditions, subjects could reach a steady state gait during each trial. Previous studies in the literature showed that the number of step to be made (one single step vs. two steps, or one step vs. multistep) and the related distance to be covered, did influence the biomechanical organization of gait initiation (e.g. Chastan et al. 2010, Dietrich et al. 1992). For short distance walks, participants may plan to brake their COM velocity very early during the initiation of gait in order to facilitate gait termination. Such early barking may “contaminate” the APA. To avoid such constraint, and also because we believe it is a natural task, participants were asked to continue walking until the end of the 5 m track after clearing the obstacle.

Chastan N, Westby GW, du Montcel ST, Do MC, Chong RK, Agid Y, Welter ML. (2010) Influence of sensory inputs and motor demands on the control of the centre of mass velocity during gait initiation in humans. Neurosci Lett. 29;469(3):400-4.

Dietrich G, Brenière Y, Do MC (1992) Organization of local anticipatory movements in single step initiation. Hum Mov Science. 13(2) : 195-210

Reply: BOS_ymax is the same as BOS_y. This point is corrected in the text.

- line 211: immediately after foot contact. Something is wrong (in the text or in the figure) since in fig. 3 the peak and slope of Rz are marked to occur before the foot contact

Reply: We agree. Thank you for revealing us this issue. The correct dating of i) the foot contact instants (vertical lines) and ii) the Rz peaks and the associated slopes are now reported in the figure 3.

Reply: Thank you again for revealing us this issue in the figure 3. The slope of the Rz trace is now correctly indicated in the figure 3 (as stated in the reply above). This slope does not correspond to the jerk since by definition, the “jerk” corresponds to the first derivative of an acceleration trace with respect to time (or third derivative of the position trace). In the present study, we recorded the position of a marker placed at the C7 spine level with the VICON system, and evaluated the slope of the acceleration trace (i.e. the “jerk”) of this marker at the swing foot contact. Now, we finally chose to remove the term “jerk” from the manuscript and, for more clarity in the terms, we computed the “slope” of both the Rz and C7 acceleration trace. This point is made clearer line 240 (paragraph “collision forces”)

- line 223: a one way repeated measures ANOVA with just two levels is more simply known as a paired t-test.

Reply: we agree. These two methods are equivalent under the present experimental conditions. One way repeated measures ANOVA is now replaced by paired-t test throughout the manuscript.

Reply: Mean values ± SE for foot lift duration, safety distance for both foot and step length are now provided in the results section.

- lines 286-288: it is not clear if the peak value of Rz and its slope are those marked by arrows in figure 3 or not, as those arrows preceded the foot contact.

Reply: As stated above, the figure 3 was thoroughly amended following your comments. Particularly, the arrows showing the Rz peak values after foot contact and the corresponding slopes are now correctly reported in this figure. We hope it is clear now.

- lines 316-317: The much larger downward-oriented centre of mass velocity at foot contact in the FFS condition (Δ=74.5%) can be considered a direct consequence of this reduced braking. Figure 6 shows that z’M at FC (i.e. CC COM velocity at FC) is larger in FFS than in RFS, but the plot do not allow to understand if COM is moving cranially (upward) or caudally (downward). By looking at figure 3, one gets even more confused, since COM at FC is moving cranially and with a lower velocity in FFS than in RFS. Moreover, considering the Braking index formula and the fact that at t₀ the z’M should be null, from figure 3 it comes out that the Braking index should be greater than 1, as V_MIN–V_FC > V_MIN. This is clearly in contrast with the results plotted in figure 6, where the average Braking index is lower than 1. Finally, the indication of Braking index in figure 3 is wrong, as it lack the 1/V_MIN factor.

Reply: we do agree with the reviewer that there was a confusion between the text, the mean values of the z’M at foot contact provided in the figure 6 (mean values all participants together) and the z’M values at foot contact shown in the trace of the figure 3 (one single trial). Thank you again for revealing us this issue.

This confusion was due to the two following points, which are now amended in the new version of the manuscript:

1) The mean values of the vertical COM velocity recorded at foot contact and which were reported in the histograms of the figure 6, were absolute values. These mean values were in fact negative (i.e. they were downwards oriented) in both the FFS and the RFS conditions. To clarify this point, negative values for this velocity are now reported in the histograms.

2) In the biomechanical traces of the figure 3, one can indeed see that the vertical COM velocity at foot contact reaches a positive value in both conditions, i.e. the velocity was oriented upward in these trials. A positive (upward) COM velocity at this instant occurred in a few trials; however, for the majority of the trials, this velocity was in fact oriented downwards, as it could now be appreciated by the negative mean values (all participants together) provided in the figure 6. This point reflects the difficulty to illustrate the mean value of an experimental variable with all participants together, with the biomechanical traces obtained from one single participant performing one single trial. These traces were however replaced by new ones which better match the general tendency.

Finally, the indication of “Braking index” in figure 3 is removed, as it lacks the 1/V_MIN factor as stressed by the reviewer.

Figures and legends

- Figure 2: why xM is not illustrated?

Reply: xM is not illustrated in this figure because no dependent variables were taken on this trace. We therefore believe that it was therefore not of much interest to report it. In addition, given the relatively high number of traces already reported in this figure, we wished not add a new one (unless special recommendation of the reviewer and/or the editor).

Reply: the sharp peak immediately following the swing foot contact is not an artefact but it reflects the collision force due to the contact of the swing foot with the force-plate. This collision force is transmitted through the whole body and it was measured at C7. This peak Rz is significantly lower in the FFS than in the RFS condition, as is the slope of this peak, which is congruent with the literature.

- Figure 3: in this figure the "Jerk" seems to start just before the heel contact. How is it possible that a collision force occurs before the strike?

Reply: We do agree that there was an error on the report of the foot contact instant in the figure 3. It is now correctly dated.

Reply. Thank you for alerting us on this point. “FC” is now used in the figures 2 and 3 instead of “TC” and “HC”.

Reply: We agree. The plots in the figures 4, 5 and 6 are now reported according to the description order followed in the text. In addition, the error bars in each of these plots are changed from +1 SD to +1 SE.

- Figure 4 and 5: please control the results of statistics. The top-left panel of figure 4 and the bottom-right panel of figure 5 report significances, while the text states the contrary.

Reply: This contradiction between the text and the figures 4 and 5 is corrected in the figure.

Minor comments

- lines 19-20: The downward vertical centre of mass velocity was less braked…, you will probably mean The downward velocity of the centre of mass was higher… or The downward fall of the centre of mass was less braked, i.e. its velocity was higher …

Reply: The sentence was reformulated for clarity. “The vertical braking of the center of mass during GI swing phase was attenuated in FFS compared to RFS, leading to greater downward center of mass velocity at foot contact in FFS.”

- line 33: for parallelism with “postural phase”, also execution phase needs ""

Reply: Done.

Reply: we agree. The sentence is replaced by the suggestion of the reviewer. Thank you.

- line 63: (cf. Yiou et al. (2017) for a recent review). For analogy to the rest of the test should be: (cf. 2, for a recent review)

Reply: OK. Change done.

- line 66: ecological, do you mean "natural"?

Reply: yes. Change done.

- line 78: gait initiating over an obstacle to be cleared … is a very long phrasing. Isn't at this point "gait initiation (GI) over a hurdle" enough?

Reply: OK. Corrected.

- line 101: for homogeneity, As in our previous study ¹¹ should be As in our previous study [11]

Reply: OK. Corrected.

- line 118: upright standards. Probably upright stands

Reply: OK. Corrected.

- line 155: a comma is needed between period and lasting

Reply: OK. A comma is added.

- line 193 and Eq. 3: all subscripts “HC” should be changed into “FC”

Reply: OK. Corrected.

- line 203: you probably mean swing, not stance

Reply: The reviewer is correct. This point is corrected.

- line 235: at the beginning of the swing phase. According to fig. 2 yM reached his peak in the late part of the swing-phase

Reply: The reviewer is right. This point is corrected.

- lines 256-257: Bars. Do you mean plots?

Reply: we were not clear with this comment of the reviewer, because we did not see the term “bars” lines 256-257, but only in the figure captions.

Reply: The reviewer is correct. The term “execution phase” is replaced by “swing phase” throughout the manuscript.

Article Menu

Printed Edition

Influence of Swing-Foot Strike Pattern on Balance Control Mechanisms during Gait Initiation over an Obstacle to Be Cleared

Further Information

Guidelines

MDPI Initiatives

Follow MDPI