Does aging affect implicit motor adaptation?

Review of the revised article “Not fleeting but lasting: Limited influence of aging on implicit adaptative motor learning and its short-term retention” by Pauline Hermans, Koen Vandevoorde, and Jean-Jacques Orban de Xivry, submitted to PCI Health and Movement Science.

The authors have provided a nice revision of their paper. Again, I think this study is very well done, the findings are interesting and make a nice contribution to the literature, and the authors have adequately addressed the majority of my questions. I only have one more remaining constructive comment below:

I appreciate the authors’ making an effort to emphasize that this is a conceptual replication. I also share their view that such conceptual replications are important, despite the fact that such studies have limitations/weaknesses. I also appreciate the addition of a paragraph in the discussion to further emphasize the conceptual replication point. However, it is worth considering the claim that is made at the end of this paragraph. It assumes that an age difference in spontaneous recovery is a finding that is “likely dependent on the experimental conditions.” By that same logic, the finding of a lack of age difference in spontaneous recovery is likely dependent on experimental conditions. I argue that a better statement would emphasize the need for further research to understand the conditions under which there is, and is not an age difference in spontaneous recovery, and to further explore the implications for those observations for the view that spontaneous recovery specifically reflects implicit learning/memory processes.

In my review of the previous version of this manuscript, I expressed my concerns regarding rationale behind the work, the specificity of the hypotheses, the description of the experimental task, and some of the analysis. These concerns have been resolved in the revised version of the manuscript and the revisions have helped to improve the quality and clarity of the work. I have no remaining comments.

The manuscript presents a comprehensive report on a behavioral study that investigates the effect of age on the contribution of implicit processes to motor adaptation and spontaneous recovery. The introduction presents relevant background information and the data-analysis approach via both frequentist and Bayesian models is an asset to the work. I also appreciate that the authors present individual data points in the figures and I believe that the discussion offers a sound, integrative interpretation of the findings. At the same time, I have several requests/questions related to clarification of the rationale behind the work, the specificity of the hypotheses, the description of the experimental task, and some of the analysis choices. I will provide more detailed comments below, which I hope will be helpful towards revising the work.

1. The introduction states that it currently is unknown whether age affects the implicit component of motor adaptation in a force-field paradigm. It would be helpful for readers if the authors could explain why it would be relevant/important to find an answer to this question. What is fundamentally different about motor adaptation in a force field versus visuomotor rotation paradigm that warrants the investigation performed in the current study? To be clear: I do agree that this is important, but just think it should be clarified more explicitly why this is the case.
2. I found the description of the timescales of different processes involved in spontaneous recovery relatively difficult to follow, and am afraid I was lost at the sentence “…which is not washed out by the few deadaptation trials whilst also rapidly forgetting”. The sentence seems to involve different theoretical constructs and grammatical structures. Would it be possible to present this in a different manner, such that the contributions of slow and fast processes are addressed separately for example?
3. When presenting the hypotheses, please clarify the directionality of the expected group differences and associations. I also recommend adding the research question / hypothesis related to working memory capacity here, as that construct is now only introduced in the methods section yet it forms an important element in the results and discussion.
4. Have the authors performed a sample size calculation or could they provide another type of justification for including 21-28 participants per group in their experiment?
5. What were the exclusion criteria for participation related to the general health and consumption habits questionnaires that were used during screening?
6. I have several questions about the description of the experimental paradigm;
   1. My assumption would be that participants could not just move the handle in the horizontal direction, but also diagonal and forward/backward, as otherwise they would be unable to reach each of the target locations. I therefore wonder is perhaps the statement in line 93 is incorrect? Or do the authors mean to say that the handle moves in 2D?
   2. Did the robotic handle return to the starting point after each trial, or did different trial start out from different hand/arm positions?
   3. To what extent can the uncued trials really be considered as uncued? Unless I misunderstood the description of the task (in which case, please revise the text as other readers may understand the text similarly), it seems that the white color of the cursor still provides information to participants about what type of trial is to come.
   4. With respect to the error-clamp phase, I think it would be helpful to briefly explain what these trials are and whether any feedback was provided. I also strongly recommend to clarify in the text that these trials were not presented on all target locations (currently this is only mentioned in the figure caption).
   5. I am not quite sure I understand what it means that there was a channel and the hand was guided through it in the error-clamp phase. Did a physical channel appear in the apparatus?
      It would be helpful to add some text-info on the phases (e.g., baseline, washout) in Figure 1, which is currently done for the error-clamp phase but not others.
7. Regarding the WM task, it would be helpful to clarify early on that it concerns a visual WM measure. My suggestion would be to describe both the task and the outcome measure in a separate WM paragraph, rather than as part of the experimental adaptation paradigm.
8. I suggest renaming the data analysis paragraph to ‘data processing and analysis’. In addition, I think the authors could consider using same order of information as in the task description; this will make the section much easier to follow.
9. Given that there are quite a larger number of analyses that are being performed, it would be helpful to clarify how they link to the hypotheses. This will also allow for a more concise, integrated presentation (instead of repeatedly saying ‘the same was done for…’). Could the authors also explain what levels are included in their multilevel correlation? I assume it includes age to account for differences between groups, but this is currently not described.
10. I missed a description of the WM correlation analyses, and would like to suggest to include age group in the model as well. Work by Seidler has shown that the association between adaptation and WM capacity is different for younger and older adults. It therefore seems reasonable to analyze the younger and older groups separately and/or include group as a factor. Not doing so could potentially have obscured any group-specific results, and the data patterns in Fig. 6 seem to possibly support this notion (though I simply base this on visual inspection, so am not sure that statistical results will concur).
11. The use of Bayesian in addition to frequentist statistics to study spontaneous recovery is an asset to the study (although the rationale behind and difference between analysis 4  and 7 were not fully clear to me). The results of the Bayesian analyses do not strongly support the conclusions in the way that PCI defines this (i.e., BF>20), yet I feel this is adequately addressed in the discussion section where the authors acknowledge that the range of effects even includes the possibility of an opposite age pattern). To further support the presented findings, the authors could also apply a Bayesian approach to analyses that go beyond spontaneous recovery, such as the age effect on adaptation and the association between adaptation and WM capacity. This would subsequently allow for interpretations such as the one currently presented in line 13 of the abstract (“both groups adapted equally well” – assuming this is indeed what the BF will support).
12. When looking at Figure 2A, it seems that there was no complete washout as the deviation does not return to zero. To what extent could this have affected the (interpretation of) the findings? It seems that this is an important point for the discussion section.
13. Line 290 may give the impression that data sets from a prior study and the present study were combined. This is a bit misleading, so I suggest to rephrase this.
14. In the discussion, the authors could address practical/clinical implications of their findings (e.g., with regards to physical therapy/rehabilitation after a fall). In addition, I recommend discussing the generalizability to longer-term memory (i.e., savings), or other forms of adaptation (e.g., rotated feedback in the manual domain, but also potential extensions to the locomotor domain).
15. Lines 447-449; It would be helpful if the authors could explain why the inter-subject variability difference between younger and older adults is a limitation. How does it affect the (boundaries of) interpretation of the present findings? I would also like to read more about how they envisage this issue being tackled. Do the authors mean that researchers should look for more homogeneous older adult groups? Or that specific methodological/statistical approaches should be developed to
16. Throughout the manuscript there are several instances were small corrections or additions to the text would be very helpful for readers. I list some examples from the introduction section here, but encourage the authors to check for other similar issues in the remainder of the manuscript;
    1. Lines 23 and 25; I suggest changing the phrasing of ‘thanks to’ to ‘via’ or ‘through’.
    2. Line 27; Could the authors please clarify what it is that is being updated? I think they may be referring to motor representations, but this is currently up to the reader to infer and would be easier to understand if it was made explicit.
    3. Line 36; What do the two sets of references mean?
    4. Line 37; This should read ‘suggests’.
    5. Line 44; Please explain what the little difference refers to; I think it refers to adaptation performance between younger and older age groups, but this should be clarified.
    6. Line 53; ‘which IS hidden’
    7. Line 67; I’m not sure how three things can be contradictory.
    8. Line 73; Start the sentence with ‘As it is known…’