Review Reports - Circadian Regulation and Pain: A Systematic Review of the Association Between Rest–Activity Rhythm and Pain-Related Outcomes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

From nearly 10,000 records identified across four databases, after the entire screening process, 7 studies remained. Seven. For all of humanity, all pain types, and decades of research on circadian rhythms.

This should make us pause before drawing clinical conclusions.

And if we look at the quality of those seven studies: one raises some concerns, four have high risk of bias, and two have very high risk of bias. In other words, not a single study in the entire review has low risk of bias. Not one.

This clearly stated in the discussion, and that honesty is appreciated. But it makes the overall tone of the article which at times seems to suggest promising clinical implications somewhat disproportionate to what the evidence actually supports.

One of the most serious problems with this review is that the studies are so different from one another that comparing them becomes almost an academic exercise. Consider:

Actigraphy duration ranges from 3 to 14 days (and the authors themselves note that 7–14 days is recommended, yet most studies fall short of that minimum).
Some use parametric models (cosinor), others non-parametric, and results point in different directions depending on the method chosen.
The metrics used vary so widely that no two studies measure exactly the same thing in the same way.
Pain assessment tools are disparate: from daily worst-pain diaries to composite scales, along with different versions of the BPI applied with different scoring criteria.

All of this means that the narrative synthesis, necessary given the impossibility of conducting a meta-analysis, is hard to interpret. The authors do their best, and they do so carefully, but readers should resist the temptation to read the results as a coherent picture. They are not.

Upon causality, all studies are observational. Four are cross-sectional, and three are longitudinal but without intervention. This means that, by design, none of them can establish causality.

When the authors write in the discussion that more robust circadian rhythms "are protective against pain," or that "improving rest-activity patterns could reduce pain," they are making a logical leap that the data do not authorize. It may be that pain deteriorates the circadian rhythm. It may be that a third variable — depression, sedentary behavior, medication — explains both. The designs of the included studies cannot distinguish between these possibilities.

The authors are aware of this and mention it briefly, but the narrative structure of the article tends to present the relationship in a single direction (circadian rhythm → pain), when bidirectionality is equally or more plausible.

Having said all of that, the review has real merits:

Methodologically, it is rigorous within what it can control. The independent dual review, the use of ROBINS-E (a tool specifically designed for observational studies, rather than the more generic ROB2), and the transparency about the protocol deviation are all positive points.

Conceptually, organizing and describing the different RAR metrics (Mesor, Amplitude, Acrophase, IS, IV, RA, I<O, R24) with clear definitions and a synthesis table is genuinely useful for researchers entering this field. The cosinor figure with its parameters is didactic and well executed.

As a map of the field, it serves its purpose: identifying what has been done, with which methods, in which populations, and where the gaps lie. That is, arguably, its most honest and valuable contribution.

There are a couple of minor points that are nevertheless worth noting:

Two of the seven included studies (Ma et al. 2014 and Chang & Lin 2014) use the same cohort of 68 patients. This is mentioned but not discussed in depth in terms of what it means for the weight of the evidence. In practice, those two studies are not independent.
The reference list contains some inconsistencies in numbering (there are duplicate references, such as [35] and [36] which appear to cite the same Ver Hoeve et al. article, and some in-text citations do not exactly match the list). These are minor errors typical of a preprint, but they should be corrected before final publication.
The authors declare having used Microsoft Copilot during writing. This is not an ethical problem if outputs were reviewed (as they state), but in the context of a systematic review that demands rigor in synthesis, it is information that readers deserve to know and weigh for themselves.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

very intersting research

please add a methods section explaing all your methodology following prisma checklist

did you record on prospero?

i appears you have 2 groups studies, is possible toperforma meta analysis?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This is the first systematic review to synthesize actigraphy-based rest–activity rhythms and pain outcomes across populations. Using rigorous PRISMA/PROSPERO methods and high inter-rater reliability, it identifies consolidated Relative Amplitude as most consistently linked to lower pain comparing to alternative metrics. Findings are contextualized by population and design, and the review prompts the neeed for standardized actigraphy protocols and wearable-based, non-pharmacological interventions to improve circadian alignment in pain management.

While I recommend this review, I invite the authors to address the following principal issues in revised version:

True comparison between parametric and non-parametric metrics remains limited by the absence of standardized time-series quality criteria, particularly for sampling density and non-wear detection thresholds. Parametric metrics are highly sensitive to data gaps, yielding substantially different results on raw versus non-wear-cleansed data, undermining cross-study comparability. The review therefore can emphasize the need for urgent consensus on minimum data quality standards to enable valid metric comparisons and reproducible research in the related fields.
The authors noticed that non-wear data were inconsistently handled, potentially biasing results. However, the heterogeneity in rest-activity rhythm–pain associations is inevitable, reflecting both the inconsistency of rest-activity rhythm assessment metrics and the multifaceted nature of pain, where distinct physiological drivers may disrupt central and peripheral clocks unevenly. The heterogeneity in rest–activity rhythm–pain associations is not merely a methodological challenge: it is an inevitable reflection of the complex, bidirectional interplay between circadian disruption and pain biology. This variability arises from two interrelated sources: first, the lack of standardized metrics for quantifying rest-activity rhythm, which capture distinct aspects of rhythmicity (e.g., amplitude, stability, timing) with above mentioned lack of standardized approach. Second, the multifaceted nature of pain itself, where distinct pathophysiological drivers such as neuroinflammation, central sensitization, or peripheral nerve dysfunction may differentially disrupt central (suprachiasmatic) and peripheral (e.g., immune, nociceptive) circadian clocks. As a result, the strength and direction of associations under review are inherently context-dependent, shaped by the underlying pain condition, disease stage, and individual circadian phenotype. Altogether, it can be empahsized that rather than viewing this heterogeneity as a limitation to be overcome, it should be interpreted as a signature of biological complexity, prompting future research toward personalized, mechanism-informed approaches rather than one-size-fits-all models.
Furthermore, only 7 small, observational studies with relatively modest sampling met criteria, with high/very high bias risk due to poor confounder control, short monitoring durations, some not covering full week, and selective exclusion. Despite rigorous analysis, as yet there is limited data generalizability, since available research focused on Western chronic pain populations.