I commend the authors for adressing the major concerns of reviewers, throughly answering the questions that were raised, and doing substantial changes to the paper. Notably, the authors included interesting new analyses on the correlation between balance control and movement efficiency changes that strengthen the message of the paper. However whith these important modifications, I still feel that some additional edits in the manuscripts are needed for readability. I will present these below:

Despite its thorough modifications, I find that the abstract could still be better structured. The early framing of deterioration VS compensation is still not easy to understand in this short form, as does the double negation "decreased [...] efficiency [...] negatively correlated with". An abstract focused on movement efficiency and with less storytelling would be easier to parse, while compensation could be mentioned only with regards to the balance results.

The more focused and streamlined introduction is appreciated. Since the authors mention that they use "effort" and "efficiency" somewhat interchangably (more on that in a comment below), I find the sentence L93-95 difficult to parse. More importantly, the end of the introduction should be amended to prepare the reader for the new analysis/results: balance is not mentioned in the introduction.

Line 125 typo: whole [B]ody

Lines 122-123 and legend of figure 1: WBR tasks still mention "a" target instead of two

Line 207: The use of the right shoulder marker for the detection of movement onset makes sense only for the whole body task, precise which marker was used for arm movements.

Lines 230 to 243: This additional paragraph about effort seems misplaced in the methods. Core concepts should be presented in the introduction (some already are), and/or recalled in the results for understandability. I'm puzzled by the sentence "we interchangably use  “energetic efficiency” and “effort minimization” in the present work". If that is the case, I would strongly suggest that the authors consistently use one term only throughout the paper for clarity. Since "effort" can have a subjective connotation, "efficiency" would seem to be the most appropriate term given the framing of the paper.

Line 247: reference [Mat, 2021] seems wrongly formatted

Figure 3: indicate again what is TA, T and NA in the legend.

Line 276:  The nomenclature "Erector Spinae D7 (ESD7)" was not correctly updated here to EST7 and elsewhere, notably in figures.

The dismissal of SOL and EST7 as postural rather than focal would deserve more explanation / justification.

LInes 300-308: Given the major concerns in the first round of review I understand the justification here, but it would now read too "defensive" for a first time reader. I feel that the main rationale should better be presented elsewhere than in the methods. Moreover, now that the authors confirmed that the machine learning was not used to select muscles of interest, I question its presentation before the main results. Could this analysis be better framed as a confirmation of the main results: antigravity muscles in whole body tasks are indeed the ones that separate subjects by age ?

Figure 6: Typo "height" instead of "eighth". On panel B I would suggest adding a line representing chance level as well as ordering the muscles by accuracy to show that antigravity muscles have the highest. The relatively high score of some non antigravity muscles (like DP) could be discussed.

I appreciate the authors' significant efforts in revising the manuscript.

Overall, they have satisfactorily addressed my comments.

However, a few minor issues remain, particularly some errors in the text.
Line 216: Please replace “analyse” with “analysis.”
Line 219: Remove the parenthesis before "Winter" and place it before the year.
Line 222: Replace “ofsset” with “offset.”
Line 230: Replace “rational” with “rationale.”
Line 234: Enclose the year in parentheses.
Line 301: Consider revising the sentence to: "Scientific literature has reported that the control of whole-body movements changes with age, while the control of arm movements does not."
Line 483: Remove the "s" from "harvests."

Regarding your response to the comment on oral consent ("The French National Ethics Committee (2019-A01558-49) approved the experiment to be conducted with oral informed consent only. Nonetheless, each participant was included in the study by a medical doctor."), it might be worth incorporating this information into the text.

The authors have addressed all my comments adequately and I have no further comments to add. 

This manuscript provides a valuable contribution by addressing the discrepancies found in prior research regarding movement efficiency in younger versus older adults. Previous studies on arm movements find that efficient control of movements is maintained in older adults, while studies on whole-body movements show degradations. However these results came from different studies, using different samples and different tasks and measurements. 
The strength of the current study is to compare both types of movements within the same sample of young and older adults and using the same type of measurement and analysis. For this, the authors adapted to whole-body movements a measure of the optimisation of movements against gravity using phasic EMG, previously used for arm movements. The study confirms the earlier observed differences between the control of arm-only movements and whole body movements in older adults. 

I however have three main concerns with the manuscript. First, since it is so central to the argument, the manuscript would benefit from a more detailed exploration of the adaptation of phasic EMG analysis to whole-body movements, taking into account potential limitations and confounds that could enrich the discussion (see point 9 below). Second, further clarification on the selection of muscles for analysis and the use of the machine learning approach is crucial to strengthen the rigor of the study (see point 10 below). Last, streamlining certain sections could improve readability and argument flow. I will provide more detailed comments and questions below, which I hope will be helpful.

Abstract / Introduction

1. The abstract could be revised for clarity and conciseness. Specifically, it should more clearly define what is meant by "behavioral compensation" and articulate the relationship between compensation and energy efficiency in the gravitational environment

2. The hypothesis presented in the abstract and introduction —that age-related compensatory processes may correspond to an adaptation process altering movement strategy— lacks clear differentiation among the terms "compensatory processes" "adaptation processes" and "changes in movement strategy". This ambiguity makes it challenging to conceive an alternative hypothesis.

3. The introduction's first two paragraphs makes a good point in explaining what compensation is, why it is important to study it and why it is difficult. However the rest of the intro could be refined, notably to present more clearly the literature on movement efficiency. In addition to point 2 above:

• The paragraph on neural mechanisms (L.72-86) could be more concise if its only point is to show that previous studies used broad measures as written L. 90.
•  L. 95 "There is no denying": informal
• The sentence L.108-110 makes the point that previously cited studies did not really study compensations and hints that studies cited in the following paragraph do. This could be be better explained and demonstrated in both paragraphs.
• The paragraph L.111-134 could benefit from a more coherent structuring to clearly present conflicts in the literature and show how the current study aims to resolve them.

Methods

4. The Methods section would benefit from additional detail to address the following points:

• The description of target placement (L. 154) should specify if it is only in front of the right shoulder, as only the right arm is used
• The spacing of targets (L. 166) needs clarification on whether two targets are used and if subjects reach with both hands.
• The methods did not specify which motion-tracking markers are used for movement segmentation for each type of movement
• L. 220: "signals were integrated...". More precisions could be added: do the authors refer to sliding-window averaging, RMS ? How does it compare/mesh with the envelope generation using low-pass filtering mentioned L. 228 ?
• L. 227-239: All steps seem to require resampled EMG, but details are missing (resampling of slow movements, how the additional delays mentioned L221-221 are taken in account for the resampling)
• L. 239: should "from the EMG trace of each fast movement" be "from the EMG trace of each pair of fast movements" since averages are made in pairs ?
• L. 264 The stated goal of the machine learning approach as described in the methods section is not as clear as in the results sections.

5. The figure 2 shows a marked decrease of the shoulder vertical position before the onset of the upward movement and after the offset of the downward movement. This raises several questions:

• What causes it ? Is it the subject bending forward before sitting up ?
• Why were onset/offset point definitions chosen so that these were ignored ?

6. In figure 3 There are identifiable bumps in the fast EMG signals visible in panel A that are not visible in panel C, notably a period at around 2/3 of the movement where there is a high variability between fast traces that is not present on the phasic traces. What causes this discrepancy ? Moreover it could be useful for comprehension to represent extracted / computed values (NA, TA, etc) in figure 3.

Results

7. The results section should specify that Figure 4 and lines 284-289 pertain to the fast movements only, if that is the case. 

8. The authors indicate that movement duration was used as a covariate in further analyses. The manuscript would benefit from an explanation of how this was done what were the outcomes of this.

9. The fact that negativities reflect the exploitation of gravity to reduce muscle efforts in arm movements is sound and well established in itself, but the jump from this to the comparison of two populations and whole body and multi-joint movements seems quite steep and doesn't explore limits and confounds. Here are a few points:

• The analysis of EMG alone seems limiting, as a difference in EMG would not be interpreted in the same way when it is accompanied by a difference in kinematics or when kinematics are the same. For example in Figure 5 for ESL1 and VL in WBR downward movements some differences in the EMG phasic traces appear to be due to timing differences rather than amplitude differences. Could those be due to different timings in the movements of different segments of the body between the groups ? How would the interpretation change if this was the case ?
• The above comment is also linked to the observation  #5 about figure 2.  If this is indeed the subject bending forward, could there be differences between the groups on this ?
• I imagine that the way gravity acts on movements would depend on the build of the person and the way muscle and fat mass are distributed in the body and limbs. The two groups could notably differ in that respect, which could be discussed.
• Negativity indices for different tasks are compared in the same statistical model. It is not immediately clear whether the normalisation process is sufficient to compare directly an index computed from two muscles in a multi-joint complex movement to an index computed from a single muscle in a single-joint simple movement.

10. L.304-318: The machine learning approach is explained more clearly here. Its goal is to confirm that that the antigravity muscles are the ones that allow the best discrimination between groups, reflecting the differences in motor strategies between groups. I find the argument and the data not convincing here, and the results raise a concern of bias in the selection of muscles:

• Some non-antigravity muscles in whole body tasks, such as DP, show classification performances that are not far off the performance of the chosen muscles (VL and ESL1)
• Some known antigravity muscles such as the soleus muscle yield comparatively poor scores here and are not selected for the main analysis. Was the machine learning approach then used for selection and not just for confirmation ? This raises a concern of potential "double dipping" if that was the case.

11. The reporting of statistical results seems incomplete. Additionally to my comment #8:

• L. 320-335: Main effects of age and tasks are not reported 
• L. 349-360: Here the mentioned models include task direction as a covariate. Was this explored in the main analysis ?

12. Fig 7 is not referred to in the text. It also does not present data from the arm task. The legend mentions different panels for data from the ESL1 and VL but this does not appear in the figure.

Discussion

13. L. 373-374: The authors mention that results suggest that the planning strategies maximise balance but this was not shown in the results L. 344

14. The similarity between paragraph L. 393-404 and the previous paragraph (same citations) should be addressed to avoid redundancy.

15. L. 426: About the non reproduction of previous results, the authors mention that the EMG-based analysis of the current paper might yield different results from the kinematic analysis from the previous study. Since motion was recorded here, the same analyses could be repeated to check this. This is linked to my comment #9.

16. The tasks proposed here differ not only in terms of arm versus whole-body involvement but also in their intrinsic nature, which could affect the control strategies employed. This could be discussed. Notably:

• Some tasks have a precision component with pointing (ARM, WBR downward D1 and D2), while others do not (STS/BTS, and WBR upward D1 and D2).
• If the WBR tasks are bimanual (see #4) this makes them also different conceptually from the ARM task.

Summary:

The paper presents a thorough investigation into age-related compensation in motor control processes, focusing on how older adults adapt their movements to maintain efficiency and functionality. The study involved twenty younger adults and twenty-four older adults, comparing their muscle activation patterns during tasks involving arm and whole-body movements. By utilizing a specific analysis method, the researchers aimed to uncover whether age-related alterations in movement efficiency reflect an adaptation process or dysfunction. The findings suggest that age-related changes in whole-body movements may be interpreted as a form of compensation for deteriorated sensorimotor components, rather than simply dysfunction. The study provides valuable insights into the intricate relationship between aging and motor control, highlighting the importance of understanding compensatory mechanisms in maintaining optimal movement patterns. However, certain aspects of the paper require further clarification, particularly concerning the definition of its central concepts and certain methodological elements. 

Overall Comments:

Central Concepts of the Study: The paper could benefit from defining and clarifying central concepts such as movement efficiency, effort, effort minimization, effort minimization downregulation or upregulation and their relationship throughout the text. For instance, the authors introduce the concept of energy efficiency in the abstract; nevertheless, it remains undefined and is not revisited in the main text. It would also enhance clarity to illustrate how the various variables measured and analyzed relate to these concepts.
Discussion on Potential Confounding Factors: Addressing potential confounding variables such as physical fitness levels and cognitive function would enhance the interpretation of the results.

Specific Comments:

Introduction:

Line 54: Consider clarifying the statement regarding the elementary concept of health.
Line 55: “despite normal age-related deterioration, compensatory processes enable older adults to remain in good health and continue to perform their daily activities comfortably.” It's important to specify that this pertains to successful aging.
Line 75: “In the sensorimotor field, following the consens that aging is associated with increased activation and increased spatial recruitment, numerous studies have attempted to establish a correlation between neural activation and behavioral performance in older adults”. Do you mean the consensus? Also, could you clarify what you mean by 'neural activation' and 'spatial recruitment'? It seems vague; are you referring to the spatial recruitment of motor units and to the neural activation of muscles?
Line 79: “This literature has not reached a consensus on the neural changes underlying compensatory mechanisms in older adults. Indeed, several studies reported a positive correlation, and as many reported no correlation or even a negative correlation”. The intended meaning is ambiguous regarding which correlation is being referenced.
Line 85: “Several reasons may explain these discrepancies”. If these reasons aren't elaborated upon, consider removing this sentence.
·        Line 88: not sure 'neuronal alterations' is the appropriate term. It sounds like you are investigating the nervous system directly. I suppose it would be better to refer to sensorimotor alterations.

·        Line 92: In the same line, not sure it's appropriate to talk about measuring neural mechanisms, given that, as you mentioned, these studies used 'broad measures'. Rather, these studies investigate behaviors through experimental paradigms that allow inference of some neural mechanisms.

·        Line 96: Still in the same vein, you write 'linking the brain to behavior.' It still sounds as if you are conducting an imagery study."

·        Line 95: Avoid 'There’s' and prefer 'There is’

Line 106: “older adults favor movement efficiency over precision to compensate for their increased energetic cost”. It's important to clarify what you mean by 'movement efficiency' (In my understanding, a movement is more efficient when achieving similar mechanical output with lower effort or neuromuscular activation). Additionally, I don't understand why you oppose movement efficiency and precision here because, in my view, these are not contradictory concepts; a precise movement can indeed be efficient. As mentioned in the overall comments, the concept of movement efficiency is central to the study and requires clearer definition and its relationship with concepts such as effort minimization and neural activation needs to be established.
Line 111: Please clarify the concepts of upregulation and downregulation of effort minimization.
Methods:

·        Line 138: Why did participants only provide oral consent instead of written consent?

·        Line 151: "vertical arm movements around the shoulder joint." Would it be more accurate to refer to shoulder flexion?

·        Line 157: Is "shoulder elevation" the appropriate term here?

·        STS/BTS task: If I am not mistaken, the rationale for using slow and fast movements is not presented.

·        Line 171: Please use "×" instead of "*".

·        Trial organization: It is not stated whether slow/fast trials and tasks were randomized.

The variables analyzed are not clearly identified, making it difficult for the reader to understand the rationale behind performing ANCOVAs in the statistics section. Further development is necessary to clarify this aspect.
Results:

Some parts appear to be more suitable for inclusion in the method section, particularly the paragraph beginning at line 304.

Line 348: I am unclear why these are categorized as exploratory analyses, as they seem to be testing your hypothesis.

Discussion:

·        Line 364: "muscle patterns" does not appear to be the appropriate term; perhaps you mean "muscle activation pattern."

·        Line 365: "Muscle command" is rather unusual and seems inappropriate given that the command is generated centrally.

·        Line 366: It could be helpful to describe this muscle marker and to rather refer to muscle activation marker of effort minimization.

·        Line 386: "Thus, arm movements equally optimized gravity effects in younger and older adults." I do not understand the intended meaning here. Do you mean that both younger and older adults optimize gravity effects to control arm movements?

·        Line 397: I would suggest placing the references at the end of the sentence.

·        Line 399: "Muscular patterns" – please refer to previous comments.

​​​