Caffeine as an Ergogenic Aid for Neuromuscular Performance: Mechanisms of Action from Brain to Motor Units

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

There is a recurring and significant error in how you name adenosine receptors. In multiple places (Section 3.2 and Section 6.1), you have written "Al receptors" using a lowercase 'L', which looks like the chemical symbol for Aluminum. This must be corrected to "A1 receptors" using the number '1'. In the "Search String" section, you used the term "peripheric adaptations" The correct scientific adjective in English is "peripheral adaptations". This might be a major shortcoming, as it yields different results.

Both Figure 1 and Figure 2 contain a spelling mistake that needs to be fixed in the image files. The word "antagonism" is misspelled as "antagoniam". Minor shortcoming

There are discrepancies between your in-text citations and the reference list. In Section 7 in the same paragraph, you discuss "Lin and colleagues [143]" but in your reference list, Lin et al. is listed as Reference 144. Y

In Section 5.1 ("Hoffman Reflex"), authors state that the use of cervicomedullary motor-evoked potentials (CMEPs) to probe spinal excitability with caffeine "remains unexplored." This is a very strong mention. It would be safer to say it is "limited" or "underexplored" as there may be existing literature using CMEPs that you have not identified.

In Section 3.1, you state that caffeine shows "higher affinity" for A2A receptors compared to A1 receptors. Caffeine is generally considered a non-selective antagonist with similar affinity for both. You should verify this statement against your cited sources

Author Response

Thank you very much for taking the time to review our manuscript. We also thank the reviewer for identifying several issues in the manuscript. All corrections are highlighted in red in the updated manuscript. We are confident that the corrections made definitely improved our manuscript.

Comment 1: "In the "Search String" section, you used the term "peripheric adaptations" The correct scientific adjective in English is "peripheral adaptations". This might be a major shortcoming, as it yields different results."

Response 1: Thank you for noticing the mistake. Corrections were made in the “Search String” section, and the term “peripheric” was corrected to “peripheral”. You can find the correction at page 3, in Section 2, now entitled “Literature Search and Review Design”, at line 103. We clarify that the actual term used in the database search was “peripheral”; therefore, the retrieved results are fully consistent with the scope of this review, and the error was limited to a misspelling. 

Comment 2: "There is a recurring and significant error in how you name adenosine receptors. In multiple places (Section 3.2 and Section 6.1), you have written "Al receptors" using a lowercase 'L', which looks like the chemical symbol for Aluminum. This must be corrected to "A1 receptors" using the number '1'"

Response 2: Thank you for pointing out. Minor shortcomings have been addressed throughout the manuscript (line 154; line 166; line 519)

Comment 3: "Both Figure 1 and Figure 2 contain a spelling mistake that needs to be fixed in the image files. The word "antagonism" is misspelled as "antagoniam". Minor shortcoming"

Response 3: Thank you for noticing the misspelling. Corrections have been made.

Comment 4: "There are discrepancies between your in-text citations and the reference list. In Section 7 in the same paragraph, you discuss "Lin and colleagues [143]" but in your reference list, Lin et al. is listed as Reference 144."

Response 4: Thank you for noticing the mistake. The reference number for the study of Lin and colleagues is now correct as number 149 (page 14, line 602).

Comment 5: In Section 5.1 ("Hoffman Reflex"), authors state that the use of cervicomedullary motor-evoked potentials (CMEPs) to probe spinal excitability with caffeine "remains unexplored." This is a very strong mention. It would be safer to say it is "limited" or "underexplored" as there may be existing literature using CMEPs that you have not identified.

Response 5: we appreciate the reviewer’s comment regarding CMEPs, and we have revised the concluding statement of the relevant paragraph to adopt a more cautious tone (page 10, Section 5.1, line 397-398).

Comment 6: In Section 3.1, you state that caffeine shows "higher affinity" for A2A receptors compared to A1 receptors. Caffeine is generally considered a non-selective antagonist with similar affinity for both. You should verify this statement against your cited sources.

Response 6: We would like to clarify our position about the affinity assumption made at page 4, in Section 3.1, line 154: in a paper by Hanajima and colleagues (2019), which is cited in support of the statement under discussion ([40] in the reference list), caffeine is described as exhibiting a slightly higher affinity for A2A receptors than for A1 receptors. In support of this, the authors refer to the study by Ciruela et al. (2006), in which the preferential affinity of caffeine for A2A receptors was demonstrated using radioligand displacement experiments. Specifically, the dissociation constant (KD) for A2A receptors was approximately one order of magnitude lower than that for A1 receptors. Furthermore, Ciruela et al. (2006) reported that the formation of A1–A2A heteromers selectively reduced the affinity of caffeine for A2A receptors, indicating that these receptor complexes represent a distinct pharmacological target. This preferential affinity was particularly evident in the striatum. Given that the primary aim of our manuscript is to highlight the role of the central nervous system in mediating the ergogenic effects of caffeine, the inclusion of the findings reported by Ciruela et al. (2006) is therefore highly pertinent to our objectives. To further make it clear for potential readers, we have added the paper from Ciruela and colleagues as reference n. 41 (page 4, Section 3.1, line 155).

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript provides a comprehensive narrative review on caffeine as an ergogenic aid for neuromuscular performance, with a strong emphasis on central and spinal mechanisms extending to motor unit behavior. The topic is timely, relevant, and well aligned with the scope of Nutrients, particularly given the increasing interest in neurophysiological determinants of performance and the use of high density electromyography. 

The conceptual framework linking adenosine receptor antagonism, corticospinal excitability, spinal modulation, and motor unit recruitment is a major strength. The sections addressing cortical excitability, perception of effort, and motor unit behavior are particularly well written and provide added value beyond prior reviews. The discussion of genetic variability, especially ADORA2A and CYP1A2 polymorphisms, is appropriate and clinically relevant.

However, as a narrative review, the manuscript would benefit from greater methodological transparency. Although a search string is reported, the inclusion and exclusion criteria are not clearly defined, and there is no description of how evidence was weighted when findings were conflicting. This limits reproducibility and may introduce selection bias. Clarifying whether this is a purely narrative or a semi systematic review would strengthen the methodological rigor.

Several mechanistic interpretations occasionally extend beyond the strength of the available evidence. In particular, the discussion of persistent inward currents and their modulation by caffeine relies on limited and heterogeneous human data, yet causal language is sometimes used. A clearer distinction between speculative hypotheses, animal data, and human experimental evidence is recommended. Similarly, the section on long term potentiation and motor learning would benefit from a more cautious tone, as the human evidence is sparse and inconsistent.

The manuscript is dense and occasionally repetitive, especially in sections describing central mechanisms of action. Condensing overlapping explanations of adenosine antagonism, dopaminergic facilitation, and arousal could improve readability without sacrificing depth. Figures are conceptually useful, but they appear to oversimplify complex neurophysiological pathways and should be explicitly framed as theoretical models rather than definitive mechanisms.

Minor issues include inconsistent use of terminology for motor unit thresholds and firing behavior, and occasional overreliance on older studies when more recent meta analyses could be emphasized more strongly in the synthesis.

Author Response

We would like to thank the reviewer for their appreciation of our work and for the thoughtful and constructive suggestions provided to improve the precision and readability of the manuscript. All corrections are highlighted in red in the updated version of the manuscript. We firmly believe that we succeeded in improving the accuracy of our manuscript.

Comment 1: However, as a narrative review, the manuscript would benefit from greater methodological transparency. Although a search string is reported, the inclusion and exclusion criteria are not clearly defined, and there is no description of how evidence was weighted when findings were conflicting. This limits reproducibility and may introduce selection bias. Clarifying whether this is a purely narrative or a semi systematic review would strengthen the methodological rigor.

Response 1: we have addressed the reviewer’s critique regarding methodological transparency by revising Section 2 (page 2) and clarifying that the manuscript is intended to be a structured narrative review rather than a systematic review. In addition, to enhance transparency, we have expanded the description of the inclusion and exclusion criteria and outlined the qualitative principles used to synthesize and weigh evidence when findings were conflicting. We believe that these additions strengthen methodological clarity while remaining consistent with the narrative scope of the review.

Comment 2: Several mechanistic interpretations occasionally extend beyond the strength of the available evidence. In particular, the discussion of persistent inward currents and their modulation by caffeine relies on limited and heterogeneous human data, yet causal language is sometimes used. A clearer distinction between speculative hypotheses, animal data, and human experimental evidence is recommended. Similarly, the section on long term potentiation and motor learning would benefit from a more cautious tone, as the human evidence is sparse and inconsistent.

Response 2: we thank the reviewer for the important observations concerning data interpretation and presentation. In response, we carefully revised the manuscript to ensure that mechanistic interpretations are consistently aligned with the strength of the available evidence. Specifically, in the sections addressing persistent inward currents (page 10, Section 5.2 – Intrinsic motor neuron properties) and their modulation by caffeine, we softened causal language and explicitly distinguished between findings derived from animal models, indirect evidence in humans, and speculative interpretations. Statements implying direct causality in humans were rephrased to better reflect the current limitations of the evidence base. Similarly, the section on long-term potentiation and motor learning (page 8, Section 4.4 – Long Term Potentiation and motor learning) was revised to adopt a more cautious tone, explicitly acknowledging the sparse and heterogeneous nature of the human literature and avoiding overgeneralization. These revisions aim to improve conceptual clarity while preserving the integrative intent of the review.

Comment 3: The manuscript is dense and occasionally repetitive, especially in sections describing central mechanisms of action. Condensing overlapping explanations of adenosine antagonism, dopaminergic facilitation, and arousal could improve readability without sacrificing depth.

Response 3: We thank the reviewer for highlighting instances of repetition within the manuscript. To address this concern, we revised the whole Section 4 (page 5 – 9) to reduce conceptual overlap and improve narrative coherence. In particular, discussions of arousal (page 6, Section 4.1), cortical excitability (page 6, Section 4.2), and perception of effort (page 7, Section 4.3) were streamlined to avoid reiteration of similar physiological concepts across subsections. Redundant descriptions of adenosine receptor antagonism and cortical activation were consolidated, with each subsection now emphasizing a distinct functional level (i.e., arousal, corticospinal transmission, or perceptual processing). In addition, the section on long-term potentiation and motor learning (page 8, Section 4.4) was reframed as a hypothesis-generating extension, clearly differentiated from the mechanisms underlying acute performance enhancement. Collectively, these revisions reduce redundancy while preserving the mechanistic coherence of the review.

Comment 4: Figures are conceptually useful, but they appear to oversimplify complex neurophysiological pathways and should be explicitly framed as theoretical models rather than definitive mechanisms.

Response 4:  According to the suggestion made, both figures were revised and the caption redesigned to make it clearer that the pathways depicted are not mechanistic interpretations but rather theoretical framings resulting from the integration of direct and speculative evidence.

Comment 5: Minor issues include inconsistent use of terminology for motor unit thresholds and firing behavior, and occasional overreliance on older studies when more recent meta analyses could be emphasized more strongly in the synthesis

Response 5: We thank the reviewer for letting us know about inconsistent use of the terminology. Accordingly, these issues were addressed throughout the manuscript (line 463; line 467; line 470; line 475; line 477; line 479; line 486; line 527; line 530; line 557; line 559; line 562; line 607; line 628).

Lastly, we would like to comment on the reviewer’s observation regarding the reliance on older studies. We agree that more recent meta-analyses could have been emphasized to a greater extent. However, as stated at the end of the Introduction (page 2, Section 1), the most recent review comprehensively addressing the specific topics of interest of our manuscript remains relatively dated (Kalmar, 2005). Consequently, in the absence of more recent reviews adequately synthesizing the literature on caffeine and voluntary movement, we relied on some dated works as foundational references while integrating newer original studies where available.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have prepared a well-written comprehensive narrative review.  I did not identify areas that required English grammar revision. 
Technically they appear to cover are relevant and pertinent topics related to caffeine consumption and performance

L57: documents?  Would studies be a more appropriate term?

Were the figures your own creation? Or does it need to be cited/attributed?

Author Response

We would like to thank the reviewer for the appreciation of our manuscript. We have adopted your suggestions which can be found highlighted in red in the updated version of the manuscript.

Comment 1: L57: documents?  Would studies be a more appropriate term?

Response 1: corrected line 57 (now line 74, page 2, Section 1)

Comment 2: Were the figures your own creation? Or does it need to be cited/attributed?

Response 2: Thank you for letting us notice the oversight. We acknowledge the missing of specification regarding the figures used. Therefore, we explicitly stated that the figures are our own creations made with a specific online software called BioRender.com (page 8 and 12).

Reviewer 4 Report

Comments and Suggestions for Authors

The present work stands out in the current scientific literature by offering a robust mechanistic synthesis that goes beyond conventional reviews on supplementation and sports performance. The authors develop a logical and well-substantiated narrative, convincingly establishing that the ergogenic effects of caffeine, at physiological doses, are predominantly mediated by the Central Nervous System. In doing so, the manuscript effectively refutes outdated theories proposing a direct action on the sarcoplasmic reticulum via ryanodine receptors in humans, given that the concentrations required for such effects would be toxic. The core argument linking adenosine receptor antagonism (A2A and A1) to modulation of corticospinal excitability and, consequently, motor unit behavior, is presented with clarity and a high level of technical depth.

The timeliness of the review is one of its strongest attributes, clearly demonstrating that the authors are operating at the forefront of knowledge in this field. The reference list is exceptionally up to date, incorporating studies published in 2024 and 2025, which lends the manuscript strong temporal relevance. The article addresses critical knowledge gaps and proposes future research directions of high heuristic value. The hypothesis regarding the influence of estrogen on adenosine receptor density regulation and its interaction with the menstrual cycle to modulate neural excitability is plausible and highlights a necessary avenue for future investigation, especially considering that most available data are derived from male participants. Furthermore, the discussion of chronic adaptations to training under caffeinated conditions situates the manuscript within contemporary debates on motor control and neuroplasticity.

Regarding the abstract, it is currently too general and lacks sufficient specificity to clearly convey the central mechanisms, methodological advances, and key implications of the review. Given that the abstract is often the first (and sometimes only) section read by potential readers, a more precise and informative abstract would substantially improve the manuscript’s clarity, accessibility, and impact. 

Despite the overall excellence of the manuscript and its significant contribution to consolidating the concept that caffeine “tunes” motor system gain via neurotransmitter modulation and spinal excitability, there are specific points that warrant the authors’ attention to ensure technical precision and argumentative coherence. These critical issues, which may be viewed as limitations of the current synthesis or areas requiring greater interpretative caution, are detailed below:

Unresolved contradictions regarding Persistent Inward Currents (PICs):

• The manuscript presents a direct conflict between studies reporting no effect of caffeine on PICs (e.g., Mackay et al.) and others suggesting changes in recruitment thresholds (e.g., Nishikawa et al.). Although the authors attempt to justify this discrepancy by citing differences in the muscles studied (Tibialis Anterior vs. Vastus Lateralis) and methodological approaches, the discussion would benefit from a deeper critical analysis of whether limitations inherent to current techniques for estimating PICs in humans might be the primary source of these inconsistencies, rather than true muscle-specific physiological differences.

Extrapolation to dynamic movements:

• The vast majority of HDsEMG evidence discussed and used to support the proposed motor unit recruitment mechanisms is derived from isometric contractions. The manuscript would benefit from being more explicit in cautioning readers that the direct transfer of these fine mechanisms to ballistic and dynamic movements, typical of real-world sports practice, remains largely theoretical due to the technical challenges associated with signal decomposition during rapid contractions.

Graphical oversimplification in Figure 2:

• While the text appropriately acknowledges that evidence for caffeine-induced increases in PIC amplitude is inconsistent and controversial, Figure 2 presents a flowchart in which an “Increase in Persistent Inward Current Amplitude” appears as a direct and linear outcome, without the necessary visual qualifiers. This creates a mismatch between the nuanced discussion in the text and the assertiveness of the figure, potentially leading to misinterpretation by readers who rely heavily on visual summaries.

Author Response

We sincerely thank the reviewer for the thorough and constructive evaluation of our manuscript and for the positive assessment of its conceptual framework, timeliness, and contribution to the field. We have carefully addressed all points raised and revised the manuscript, accordingly, as detailed below. All adjustments can be found highlighted in red in the updated version of the manuscript.  Overall, these revisions were implemented to enhance technical precision, reduce interpretative overreach, and ensure consistency between text, figures, and conclusions. We believe these changes have substantially strengthened the manuscript and improved its clarity and balance.

Comment 1: Regarding the abstract, it is currently too general and lacks sufficient specificity to clearly convey the central mechanisms, methodological advances, and key implications of the review. Given that the abstract is often the first (and sometimes only) section read by potential readers, a more precise and informative abstract would substantially improve the manuscript’s clarity, accessibility, and impact. 

Response 1: Thank you for raising concerns about the way the abstract was presented. Following your suggestion, the abstract was revised to improve specificity and clarity. It now explicitly highlights the central mechanistic focus of the review, the predominance of central nervous system–mediated effects of caffeine at physiological doses, the key implications for neuromuscular performance and future research.

Comment 2:  Unresolved contradictions regarding Persistent Inward Currents (PICs):

• The manuscript presents a direct conflict between studies reporting no effect of caffeine on PICs (e.g., Mackay et al.) and others suggesting changes in recruitment thresholds (e.g., Nishikawa et al.). Although the authors attempt to justify this discrepancy by citing differences in the muscles studied (Tibialis Anterior vs. Vastus Lateralis) and methodological approaches, the discussion would benefit from a deeper critical analysis of whether limitations inherent to current techniques for estimating PICs in humans might be the primary source of these inconsistencies, rather than true muscle-specific physiological differences.

Response 2: To address the concern regarding the contradictions in the literature, we substantially revised the section on PICs (page 10, Section 5.2) to include a deeper critical appraisal of the methodological approaches used to estimate PIC amplitude in humans. We now explicitly discuss the indirect and assumption-dependent nature of EMG-based estimates, the potential violation of these assumptions following caffeine ingestion, and the limited sensitivity of current techniques. This reframing emphasizes that inconsistencies across studies may arise from methodological constraints rather than true muscle-specific physiological differences, and causal language has been further attenuated throughout the section.

Comment 3: Extrapolation to dynamic movements:

• The vast majority of HDsEMG evidence discussed and used to support the proposed motor unit recruitment mechanisms is derived from isometric contractions. The manuscript would benefit from being more explicit in cautioning readers that the direct transfer of these fine mechanisms to ballistic and dynamic movements, typical of real-world sports practice, remains largely theoretical due to the technical challenges associated with signal decomposition during rapid contractions.

Response 3: In response to the reviewer’s comment, we strengthened the cautionary language regarding the transferability of motor unit–level mechanisms derived from isometric HDsEMG studies to dynamic and ballistic movements. The revised text (page 11, Section 6, line 486-492) now explicitly states that such extrapolations remain largely theoretical, given current technical limitations in signal decomposition during rapid and non-isometric contractions.

Comment 4: Graphical oversimplification in Figure 2:

• While the text appropriately acknowledges that evidence for caffeine-induced increases in PIC amplitude is inconsistent and controversial, Figure 2 presents a flowchart in which an “Increase in Persistent Inward Current Amplitude” appears as a direct and linear outcome, without the necessary visual qualifiers. This creates a mismatch between the nuanced discussion in the text and the assertiveness of the figure, potentially leading to misinterpretation by readers who rely heavily on visual summaries.

Response 4: Figure 2 and its caption were revised to better reflect the uncertainty and heterogeneity of the available evidence regarding caffeine-induced modulation of PICs (page 12, line 494). Visual qualifiers were added, the solid arrow has been switched with a dashed one to highlight the theoretical nature of the diagram, and the figure is now explicitly presented as a simplified theoretical representation. The caption clearly states that PIC-related effects are inferred from indirect measures and should be interpreted with caution, thereby aligning the graphical content with the nuanced discussion in the text.