Velocity-Dependent Eccentric Knee-Flexion Isokinetic Assessment in Elite Professional Soccer Players: Reliability, Inter-Limb Asymmetry and Mechanical Characteristics

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript addresses a relevant topic in sports science and provides useful descriptive data on eccentric isokinetic performance in elite soccer players. However, it is primarily descriptive, with limited novelty and several methodological concerns, particularly regarding statistical handling, clarity of variable definitions, and overinterpretation of findings in relation to injury prevention.

Specific commentaries bellow:

Page 1, Line 16–18: The objective combines descriptive and comparative aims; clarify primary vs. secondary outcomes to improve focus and hypothesis clarity.

Page 1, Line 19–23: The sample is described as “healthy professional soccer players,” but no justification of sample size or power analysis is provided; include an a priori sample size calculation.

Page 1, Line 21–23: The inclusion of multiple dependent variables (PT, PTM, APT, TW, H:Q, RTD) increases multiplicity risk; clarify whether corrections for multiple comparisons were applied.

Page 1, Line 24–27: The results lack exact p-values and confidence intervals; report them consistently for transparency.

Page 1, Line 28–30: The conclusion overgeneralizes the findings; soften claims regarding “comprehensive neuromuscular profiling” given the cross-sectional design.

Page 2, Line 41–43: The introduction mentions injury prevention but does not clearly link study variables to injury outcomes; strengthen the rationale.

Page 2, Line 61–69: The discussion of injury prediction is presented, but the study does not assess injury risk; clarify this distinction earlier.

Page 2, Line 74–80: The methodological rationale for selecting 60°·s⁻¹ and 180°·s⁻¹ is insufficient; justify these velocities with stronger evidence.

Page 2, Line 82–85: The research gap is vaguely stated; explicitly define what is unknown and how this study addresses it.

Page 3, Line 94–100: The study design is cross-sectional but includes reliability analysis; clarify whether this is intra-session reliability only.

Page 3, Line 98–100: Using five repetitions for reliability assessment may be insufficient; justify this choice with prior literature.

Page 3, Line 102–106: Exclusion criteria are listed, but inclusion criteria are not clearly defined; specify them explicitly.

Page 3, Line 108–110: Ethical approval date (2017) appears outdated relative to publication year; clarify if data collection occurred earlier.

Page 3, Line 113–121: The dynamometer calibration procedure is not described; include calibration details to improve reproducibility.

Page 3, Line 121–123: Range of motion is vaguely reported (“90–100°”); specify exact ROM used.

Page 3, Line 125–127: The testing order is fixed; discuss potential order or fatigue effects and whether randomization was considered.

Page 3, Line 128–130: Criteria for removing “invalid trials” are unclear; define objective thresholds.

Page 3, Line 131–137: The definition of MPT is unclear (“mean of the five highest peak torque values”); clarify whether this differs from PT.

Page 3, Line 145–147: RTD calculation is limited to 0–50 ms; justify this time window and discuss sensitivity limitations.

Page 4, Line 150–152: Normality testing is mentioned, but results are not reported; indicate whether assumptions were met.

Page 4, Line 152–154: Use of multiple paired t-tests increases type I error; consider multivariate approaches or correction methods.

Page 4, Line 155–158: Repeated-measures ANOVA is used across repetitions, but sphericity assumptions are not addressed.

Page 4, Line 159–161: ICC model (2,1) is stated, but justification for this choice is missing; clarify model selection.

Page 4, Line 162–169: Equations for SEM and MDC are presented but formatting is inconsistent; improve readability.

Page 4, Line 171–172: The alpha level is set at 0.05 without correction for multiple comparisons; address this limitation.

Page 4, Table 1: There are formatting inconsistencies (e.g., “0.85.1±10.9”); correct typographical errors.

Page 4, Table 1: Confidence intervals for effect sizes are inconsistently formatted; standardize presentation.

Page 5, Table 1: Units and abbreviations are excessive and reduce readability; simplify or move to supplementary material.

Page 5, Line 196–198: Effect sizes are described as significant despite being small; interpret practical relevance more cautiously.

Page 5, Line 199–201: The shift in APT is described but lacks mechanistic explanation; expand interpretation.

Page 6, Figure 1: The figure lacks clear labeling of statistical differences between limbs; improve annotation clarity.

Page 6, Figure 2: Dual-axis representation may confuse interpretation; consider separating variables into different figures.

Page 7, Line 215–218: Stability across repetitions is reported, but no statistical values are provided; include ANOVA results.

Page 7, Table 2: Reliability metrics are only reported for PT; justify why other variables were excluded.

Page 7, Line 220–227: SEM values are very low; discuss whether this reflects true precision or methodological bias.

Page 8, Line 245–248: The interpretation of symmetry may overlook individual variability; include variability analysis.

Page 8, Line 260–263: The explanation for dominant limb TW difference is speculative; acknowledge uncertainty more explicitly.

Page 8, Line 269–272: Normative comparisons are made without describing reference populations; clarify comparability.

Page 9, Line 297–301: The force–velocity relationship is mentioned but not quantitatively analyzed; consider including regression analysis.

Page 9, Line 303–307: The interpretation of APT shifts as injury-related is speculative without injury data; moderate claims.

Page 9, Line 309–313: Higher TW at higher velocity contradicts lower PT; provide deeper physiological explanation.

Page 9, Line 322–326: The absence of fatigue effects is claimed but not directly measured; rephrase cautiously.

Page 10, Line 340–344: Limitations section is incomplete; include statistical and methodological limitations (e.g., multiple testing, lack of randomization).

Page 10, Line 346–348: The exclusion of higher velocities (≥300°·s⁻¹) is noted but not justified in design; clarify rationale.

Page 10, Line 350–353: The absence of EMG or imaging limits interpretation; this should be emphasized earlier as a limitation.

Page 10, Line 355–360: Conclusions restate results but extend to injury prevention without evidence; reduce overinterpretation.

Page 10, Line 365–368: Recommendation of multiple velocities is reasonable but should be framed as preliminary.

Page 10, Line 371–374: Practical applications suggest clinical relevance without validation; moderate strength of recommendations.

Page 10, Line 385–387: MDC interpretation is appropriate but lacks context of longitudinal monitoring; clarify applicability.

Author Response

Comment 1 : The manuscript addresses a relevant topic in sports science and provides useful descriptive data on eccentric isokinetic performance in elite soccer players. However, it is primarily descriptive, with limited novelty and several methodological concerns, particularly regarding statistical handling, clarity of variable definitions, and overinterpretation of findings in relation to injury prevention.

 

Response 1 : Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

 

Specific commentaries bellow:

 

Comment 2 : Page 1, Line 16–18: The objective combines descriptive and comparative aims; clarify primary vs. secondary outcomes to improve focus and hypothesis clarity.

 

Response 2 : Thank you for your comment. We have amended it accordingly, and it now reads: "The primary aim of this study was to compare eccentric knee flexion isokinetic performance at two commonly used angular velocities (60°·s⁻1 and 180°·s⁻1) between dominant and non-dominant limbs in healthy professional soccer players, through the analysis of peak torque, mean peak torque, angle of peak torque, total work, and rate-of-torque-development–related variables. The secondary aim was to describe concentric knee extension and knee flexion strength variables assessed at 60°·s⁻1."

 

Comment 3 : Page 1, Line 19–23: The sample is described as “healthy professional soccer players,” but no justification of sample size or power analysis is provided; include an a priori sample size calculation.

 

Response 3 : Thanks to the reviewer for the recommendation. This is a specific, naturally occurring group measured throughout the 2023–2024, 2024–2025, and early 2025–2026 seasons, for which exclusion criteria were established for data analysis. Therefore, all players who did not meet the exclusion criteria were included. Even so, we have introduced the seasons in which they were evaluated within the text and it can be read like this: "Forty male professional soccer players (23.7 ± 3.5 years) were recruited from a Swiss first-division soccer club competing in the UEFA Conference League during 2023-2024, 2024-2025 and 2025-2026 seasons. The term “healthy professional soccer players” was used based on the exclusion criteria: Lack of full availability for regular participation in organized training and matches; a lower-limb injury in last 6 weeks; history of lower-limb surgery in the previous 12 months; current pain or symptoms preventing maximal testing and any medical conditions contraindicating high-intensity neuromuscular testing. »

Regarding power analysis,iIt is well documented by several articles that post hoc power calculations could not be useful (Althouse, 2020; Goodman & Berlin, 1994; Hoenig & Heisey, 2001). Based on this, and on the fact that we have tables within the results with a lot of information (change in mean, effect sizes (ES) and P value). We think that we will overload the table with more information, and it will not provide substantial information in addition to the sum of ES + P value. However, if the reviewer considers that we have to include it, please confirm it and we will do it. Thanks again.

 

Comment 4 : Page 1, Line 21–23: The inclusion of multiple dependent variables (PT, PTM, APT, TW, H:Q, RTD) increases multiplicity risk; clarify whether corrections for multiple comparisons were applied.

 

Response 4 : Thank you for raising this relevant point. We acknowledge that analyzing multiple dependent variables may increase the risk of Type I error. However, no formal correction for multiple comparisons was applied, as the selected variables (PT, PTM, APT, TW, H:Q, RTD) were conducted separately for each variable. Maybe, in paired-samples t-tests was no clear in the text, so we have amended it accordingly, and now us read: "Differences between the dominant and non-dominant limbs were analyzed using paired-samples t-tests for each concentric and eccentric variables."

 

Comment 5 : Page 1, Line 24–27: The results lack exact p-values and confidence intervals; report them consistently for transparency.

 

Response 5 : : Thank you for your comment, we have amended it accordingly and and it now reads « No significant inter-limb differences were observed for most variables (trivial–small effect sizes), except for higher eccentric TW at 180°·s⁻¹ in the dominant limb (p = 0.009). Eccentric PT and PTM decreased at higher velocities in both dominant (p = 0.002 and p<0.001, respectively) and non-dominant limbs (p = 0.008 and p<0.001, respectively), while APT shifted toward more flexed knee angles (p<0.001) ».

 

Comment 6 : Page 1, Line 28–30: The conclusion overgeneralizes the findings; soften claims regarding “comprehensive neuromuscular profiling” given the cross-sectional design.

 

Response 6 : Thank you for your comment. We have amended it accordingly, and and it now reads: "Eccentric knee flexion assessment at 60°·s⁻1 and 180°·s⁻1 angular velocities provided differents results in PT, PTM and APT for the same group of players,  supporting the use of more than one eccentric test for obtained information about this variables in elite soccer. "

 

Comment 7 : Page 2, Line 41–43: The introduction mentions injury prevention but does not clearly link study variables to injury outcomes; strengthen the rationale.

 

Response 7 : Thank you for your comment. In this paragraph, we refer to all the states in which this type of test is appropriate for an athlete. On the one hand, there is the healthy athlete who is training, for whom this test provides information on potential injury prevention. On the other hand, there is the athlete nearing the end of their recovery, practically healthy, who wants to return to competition safely. This test is commonly used in these scenarios, so we are introducing the reader who falls into one of these categories, as this study will offer valuable information to help them make a more informed decision.

 

Comment 8 : Page 2, Line 61–69: The discussion of injury prediction is presented, but the study does not assess injury risk; clarify this distinction earlier.

 

Response 8 : Thank you for your consideration. We indicate that the values ​​we will analyze in our study are commonly used as predictors of muscle injury, thus introducing the reader to the research problem. If these values ​​serve as a reference for establishing potential injury risks, and as we will see, the values ​​differ depending on the angular velocity at which they are measured, we will not only have to obtain these reference values ​​to avoid risk, but we will also have to reproduce the same angular velocity used to measure them.

 

Comment 9 : Page 2, Line 74–80: The methodological rationale for selecting 60°·s⁻¹ and 180°·s⁻¹ is insufficient; justify these velocities with stronger evidence.

 

Response 9 : Thank you for your consideration. Isokinetic strength assessments are commonly performed at angular velocities such as 60°·s⁻¹ and 180°·s⁻¹, which are widely adopted in both clinical and research settings to evaluate maximal strength and more functional, velocity-dependent performance, respectively. These velocities are also frequently used for both concentric and eccentric testing protocols. We have added a reference to Dvir's work (2004), which also confirms this.

 

Comment 10 : Page 2, Line 82–85: The research gap is vaguely stated; explicitly define what is unknown and how this study addresses it.

 

Response 10 : Thank you for your comment. We have amended it accordingly, and now us read: "Despite the widespread use of isokinetic testing to assess hamstring function in soccer players, important methodological aspects remain insufficiently understood. In particular, it is unclear whether eccentric hamstring performance can be adequately characterized using a single testing velocity or whether assessments at different angular velocities provide complementary and non-redundant information. This issue is especially relevant given current recommendations encouraging multi-variable and multi-velocity profiling of neuromuscular function, while the extent to which different velocities capture distinct aspects of eccentric function remains unknown in elite soccer players. Furthermore, there is limited evidence comparing how multiple outcome variables (e.g., peak torque, an-gle-specific measures, and rate-of-torque-development–related metrics) behave across velocities and between limbs, and whether they provide overlapping or unique infor-mation. "

 

Comment 11 : Page 3, Line 94–100: The study design is cross-sectional but includes reliability analysis; clarify whether this is intra-session reliability only.

 

Response 11 : Thank you for your comment. We have amended it accordingly, and now us read: "A cross-sectional observational study design was used to examine concentric and eccentric isokinetic knee strength characteristics in professional soccer players. All participants completed a single testing session during which concentric knee extension and flexion were assessed at 60°·s⁻1, followed by eccentric knee flexion testing at 60°·s⁻1 and 180°·s⁻1. Five maximal repetitions were performed for each testing condition to allow the assessment of peak torque consistency, inter-limb comparisons, and the development of intra-session reliability indices. "

 

Comment 12 : Page 3, Line 98–100: Using five repetitions for reliability assessment may be insufficient; justify this choice with prior literature.

 

Response 12 : Thank you for your comment. We acknowledge that the number of repetitions performed during isokinetic testing may influence the reliability and stability of the measurements. The use of five maximal repetitions in the present study was based on previous literature and methodological recommendations in isokinetic testing, particularly in soccer populations. Several studies included in this field have employed a similar range of repetitions (typically between 3 and 5) to obtain reliable estimates of peak torque and related variables while minimizing fatigue-related declines in performance. These protocols are consistent with current practice in professional football settings, where a limited number of maximal repetitions is preferred to balance measurement reliability and neuromuscular fatigue. In addition, methodological references indicate that a small number of maximal repetitions is sufficient to achieve stable peak torque values, particularly at commonly used angular velocities, and that increasing the number of repetitions may introduce fatigue and compromise the validity of the assessment. Recent consensus-based guidelines in football populations also emphasize the importance of standardized protocols without recommending a substantially higher number of repetitions. We have amended it accordingly, and now us read: "The selection of five repetitions represents a compromise between ensuring measurement reliability and minimizing fatigue, and is consistent with established protocols in the literature. This rationale has now been clarified in the revised manuscript [1,3,8].

 

Comment 13 : Page 3, Line 102–106: Exclusion criteria are listed, but inclusion criteria are not clearly defined; specify them explicitly.

 

Response 13 : Thank you for your comment. The inclusion criteria are described above when we say "Forty male professional soccer players were recruited from a Swiss first-division soccer club competing in the UEFA Conference League during 2023-2024, 2024-2025 and 2025-2026 seasons."

 

Comment 14 : Page 3, Line 108–110: Ethical approval date (2017) appears outdated relative to publication year; clarify if data collection occurred earlier.

 

Response 14 : Thank you for your comment. It's true that the dates may seem a bit outdated, but it's a protocol endorsed by the two prestigious hospitals that the Andalusian Regional Government uses as a reference for validating this type of research. We could have obtained the one from our home university, since these are very common tests in football that are included by default in the assessment of a player's physical condition. However, we wanted an external center to endorse this type of intervention from an ethical standpoint, and therefore, in our application, we submitted for evaluation all the tests we typically use in the assessment of professional footballers.

 

Comment 15 : Page 3, Line 113–121: The dynamometer calibration procedure is not described; include calibration details to improve reproducibility.

 

Response 15 : Comment acknowledges. In line with several studies using isokinetic devices, no dynamometer calibration procedure was reported. In our case, prior to initiating the isokinetic protocol, the participant’s leg was weighted for gravitational error torque in accordance with the manufacturer’s procedures to account for its effect on torque. In addition, reproducibility showed very good values. This information has been included in the Methods section.

 

Comment 16 : Page 3, Line 121–123: Range of motion is vaguely reported (“90–100°”); specify exact ROM used.

 

Response 16 : Thank you for your comment. The manuscript has been revised to include the following clarification:

“The dynamometer axis was carefully aligned with the knee joint center. The range of motion was standardized to 80°, spanning from 90° to 10° of knee flexion, with 0° defined as full knee extension.”

 

 

Comment 17 : Page 3, Line 125–127: The testing order is fixed; discuss potential order or fatigue effects and whether randomization was considered.

 

Response 17 : Thank you for the comment. The order of the tests was fixed. Since this is an assessment conducted with elite professional players, it is always performed in the same way and in the same order to allow comparisons throughout the season with healthy players or with players returning from injury. As reflected in the results of the present study, this protocol shows optimal levels of reproducibility and reliability, and the selected number of repetitions is not significantly influenced by muscle fatigue, as there are no statistical differences in the different variables across the five repetitions. If we wanted to analyze a decline in force production (fatigue), based on our results we would need to increase the number of repetitions. This could be interesting to explore in a future study with this objective.

 

 

Comment 18 : Page 3, Line 128–130: Criteria for removing “invalid trials” are unclear; define objective thresholds.

 

Response 18 : Trials were considered invalid and excluded from analysis if any of the following occurred: (i) failure to reach or maintain the preset angular velocity, (ii) incomplete execution of the predefined range of motion, (iii) evident submaximal effort or inconsistent torque production, (iv) technical or recording errors, or (v) interruption due to pain or discomfort. This information has been included in the Methods section.

 

 

Comment 19 : Page 3, Line 131–137: The definition of MPT is unclear (“mean of the five highest peak torque values”); clarify whether this differs from PT.

 

Response 19 : The difference lies in taking the highest value of the five or the average of the five values.

 

Comment 20 : Page 3, Line 145–147: RTD calculation is limited to 0–50 ms; justify this time window and discuss sensitivity limitations.

 

Response 20 : Thank you very much for the comment. We appreciate it, as the information was incomplete. Changes were made to the Methods section of the manuscript, and it now reads as follows: “Peak torques were further extracted for the time intervals 0–50 ms, 50–100 ms and 100–200 ms after onset, and Rate of Torque Development (RTD), calculated as Δtorque/Δtime was measured as the slope of the MVC torque-time-curve in the time intervals 0–50 ms, 0–100 ms and 0–200 ms”.

 

 

Comment 21 : Page 4, Line 150–152: Normality testing is mentioned, but results are not reported; indicate whether assumptions were met.

 

Response 21 : Thank you for your comment. It is understood that when parametric tests are applied, it is because the assumptions for their use were met. I don't usually see the results of the normality test in the articles.

 

Comment 22 : Page 4, Line 152–154: Use of multiple paired t-tests increases type I error; consider multivariate approaches or correction methods.

 

Response 22 : Thank you for raising this relevant point. We acknowledge that analyzing multiple dependent variables may increase the risk of Type I error. However, no formal correction for multiple comparisons was applied, as the selected variables (PT, PTM, APT, TW, H:Q, RTD) were conducted separately for each variable. Maybe, in paired-samples t-tests was no clear in the text, so we have amended it accordingly, and now us read: "Differences between the dominant and non-dominant limbs were analyzed using paired-samples t-tests for each concentric and eccentric variables."

 

Comment 23 : Page 4, Line 155–158: Repeated-measures ANOVA is used across repetitions, but sphericity assumptions are not addressed.

 

Response 23 : Thank you for this important comment. In the present study, sphericity was assessed using Mauchly’s test. When the assumption was violated, Greenhouse–Geisser corrections were applied to adjust the degrees of freedom. We have amended it accordingly, and now us read: To examine potential differences across the five repetitions, a one-way repeated-measures analysis of variance (ANOVA) was conducted separately for each variable, limb, and testing condition. The assumption of sphericity was evaluated using Mauchly’s test, and when violated, the Greenhouse–Geisser correction was applied to adjust the degrees of freedom.  When significant main effects were detected, post hoc pairwise comparisons with Bonferroni adjustment were applied.

 

Comment 24 : Page 4, Line 159–161: ICC model (2,1) is stated, but justification for this choice is missing; clarify model selection.

 

Response 24 : Thank you for this important comment. The ICC (2,1) model was selected based on the recommendations of Koo and Li, as it corresponds to a two-way random-effects model assessing absolute agreement for single measurements. This model is appropriate when both subjects and measurements (i.e., repetitions) are considered random effects and when the aim is to generalize the reliability findings beyond the specific sample and testing occasions included in the study. In the present study, repetitions were not treated as fixed conditions but rather as repeated measurements representative of a broader testing context. Therefore, the use of ICC (2,1) allows for a more generalizable estimation of reliability.

 

Comment 25 : Page 4, Line 162–169: Equations for SEM and MDC are presented but formatting is inconsistent; improve readability.

Response 25 : Thank you for your comment, we have amended it accordingly.

 

Comment 26 : Page 4, Line 171–172: The alpha level is set at 0.05 without correction for multiple comparisons; address this limitation.

 

Response 26 : Thank you for this comment. We acknowledge that the statement “the level of statistical significance was set at p < 0.05 for all analyses” may have caused confusion in the context of reliability analyses. For the ICC, SEM, and MDC₉₅ calculations, significance testing and multiple-comparison corrections are not applicable, as these metrics are descriptive estimates of reliability and measurement error rather than inferential tests. To improve clarity, the revised manuscript now explicitly states that p < 0.05 refers only to inferential statistical analyses, and the reliability statistics are reported independently. We have amended it accordingly, and now us read: " MDC₉₅ values were reported in absolute units (Nm) to facilitate clinical interpretation. For all inferential analyses (e.g., repeated-measures ANOVA), the level of statistical significance was set at p < 0.05. Reliability metrics (ICC, SEM, MDC₉₅) are descriptive and not subject to hypothesis testing or multiple-comparison correction".

 

Comment 27 : Page 4, Table 1: There are formatting inconsistencies (e.g., “0.85.1±10.9”); correct typographical errors.

 

Response 27 : Thank you for your comment, we do it.

 

 

Comment 28 : Page 4, Table 1: Confidence intervals for effect sizes are inconsistently formatted; standardize presentation.

 

Response 28 : Thank you for your comment, we do it.

 

 

Comment 29 : Page 5, Table 1: Units and abbreviations are excessive and reduce readability; simplify or move to supplementary material.

 

Response 29 : Thank you for your comment : It is true that the number of variables may seem excessive, but they are all those typically described in isokinetic assessment. We represent them with the commonly used abbreviations: Variable, the movement in which it is measured, and its units of measurement.

 

Comment 30 : Page 5, Line 196–198: Effect sizes are described as significant despite being small; interpret practical relevance more cautiously.

 

Response 30 : Thank you for your comment. We do this as follows. In this section, we simply present the results honestly, but in the discussion section, we interpret these data with appropriate caution whe we said: “Effect sizes ranging from trivial to small indicated limited practical relevance. »

 

Comment 31 : Page 5, Line 199–201: The shift in APT is described but lacks mechanistic explanation; expand interpretation.

 

Response 31 : Thank you for your comment. We do this as follows. In this section, we simply present the results honestly, but in the discussion section, we interpret these data with appropriate caution whe we said: “A key finding of the present study was the systematic shift of the APT toward more flexed knee positions at the higher angular velocity (180°·s⁻1). This behavior has been consist-ently described in the literature and is thought to reflect alterations in muscle–tendon interaction and neuromuscular control as contraction speed increases [15, 17]. In this context, the reduced APT observed at 180°·s⁻1 may indicate a diminished capacity to produce peak force at longer muscle lengths, a factor of particular relevance for injury prevention screening [11, 18]. »

 

 

Comment 32 : Page 6, Figure 1: The figure lacks clear labeling of statistical differences between limbs; improve annotation clarity.

 

Response 32 : Thank you for your comment. Since all comparisons are significant, we have decided to use a single significance symbol (*).

 

Comment 33 : Page 6, Figure 2: Dual-axis representation may confuse interpretation; consider separating variables into different figures.

 

Response 33 : Thank you for your comment. Since there are only two variables, we wanted to avoid taking up too much space in the document to show something that is very visible. This strategy is commonly used in the scientific field.

 

Comment 34 : Page 7, Line 215–218: Stability across repetitions is reported, but no statistical values are provided; include ANOVA results.

 

Response 34 : Thank you for your comment. There were no significant differences between any repetitions.

 

Comment 35 : Page 7, Table 2: Reliability metrics are only reported for PT; justify why other variables were excluded.

 

Response 35 : Thank you for this comment. In this study, reliability analyses were performed only for peak torque (PT) because it is the most widely used and clinically relevant indicator of hamstring and quadriceps function in professional soccer players.

 

Comment 36 : Page 7, Line 220–227: SEM values are very low; discuss whether this reflects true precision or methodological bias.

 

Response 36 : Thank you for this insightful comment. The relatively low SEM values observed in our study likely reflect the combination of a standardized testing protocol, familiarization of participants with the isokinetic procedures, and the use of a homogeneous sample of professional soccer players, which reduces between-subject variability. Such low SEM values have also been reported in similar populations under controlled conditions. Nonetheless, we acknowledge that methodological factors, such as the limited number of repetitions (five per condition) and the highly trained sample, may have contributed to reduced variability and therefore lower SEM. This limitation has now been explicitly discussed in the revised manuscript as: "Additionally, the observed standard error of measurement (SEM) values were relatively low. While this likely reflects the standardized testing protocol, participant familiarization, and the use of a homogeneous sample of professional soccer players, it is also possible that methodological factors—such as the limited number of repetitions per condition—contributed to reduced variability. Therefore, the reported SEM values may represent both true measurement precision and potential methodological bias. Future studies should consider evaluating reliability across multiple variables and larger, more heterogeneous samples to better characterize the generalizability of these findings."

 

Comment 37 : Page 8, Line 245–248: The interpretation of symmetry may overlook individual variability; include variability analysis.

 

Response 37 : Thank you for this comment. We acknowledge that individual variability may provide additional insight; however, the primary aim of our study was to evaluate group-level differences and commonly used clinical indicators of symmetry, consistent with established practice in professional soccer research. The symmetry index is intended as a standardized metric for clinical interpretation rather than as a tool for exploring individual variability. Including additional analyses of individual differences would not alter the main conclusions and is beyond the scope of this study.

 

Comment 38 : Page 8, Line 260–263: The explanation for dominant limb TW difference is speculative; acknowledge uncertainty more explicitly.

 

Response 38 : Thank you for your comment, we have amended it accordingly and now we can read as : The only variable showing a significant inter-limb difference was total work during eccentric knee flexion testing at 180°·s⁻1, with the dominant limb producing larger amounts (moderate effect size). This isolated finding may reflect subtle neuromuscular or coordination-related adaptations associated with preferred kicking limb usage, as sug-gested in earlier studies examining eccentric hamstring loading during soccer-specific actions [11, 13]. However, given the preservation of symmetry in peak torque and H:Q ratios, this difference should be interpreted cautiously. Overall, the present findings reinforce the prevailing evidence that bilateral symmetry is typical in professional soccer players, but more studies are necessary to analyze the reason for a possible asymmetry in the total work variable.

 

Comment 39 : Page 8, Line 269–272: Normative comparisons are made without describing reference populations; clarify comparability.

 

Response 39 : Thank you for your comment, we have amended it accordingly and now we can read as :  Concentric knee extension PT values (~210–215 Nm), knee flexion PT (~150 Nm) at 60°·s⁻¹, and eccentric knee flexion PT at 60°·s⁻¹ and 180°·s⁻¹ (~170–180 Nm) fell within established normative ranges for athletes in general and footballers in particular

 

Comment 40 : Page 9, Line 297–301: The force–velocity relationship is mentioned but not quantitatively analyzed; consider including regression analysis.

 

Response 40 : Thank you for your comment, but it is beyond the scope of this study.

 

Comment 41 : Page 9, Line 303–307: The interpretation of APT shifts as injury-related is speculative without injury data; moderate claims.

 

Comment 41 : Thank you for your comment. This interpretation of ATP has been delete.

 

Comment 42 : Page 9, Line 309–313: Higher TW at higher velocity contradicts lower PT; provide deeper physiological explanation.

 

Response 42 : Thank you for comment. The observation that total work (TW) can be higher at higher angular velocities despite lower peak torque (PT) is a well-documented phenomenon in muscle physiology. While PT represents the maximal torque at a single point in the range of motion, TW reflects the integrated torque across the entire movement, taking into account the speed and range of motion over multiple repetitions. At higher velocities, although the instantaneous torque is reduced due to the force–velocity relationship of skeletal muscle, the limb covers a greater angular displacement in the same time period and more repetitions may be completed within a given protocol, resulting in higher cumulative work. Therefore, TW and PT measure related but distinct aspects of muscular performance: PT captures maximal instantaneous capacity, while TW represents the total energetic output over the movement.

 

Comment 43 : Page 9, Line 322–326: The absence of fatigue effects is claimed but not directly measured; rephrase cautiously.

 

Response 43 : Thank you for your comment, we have amended it accordingly and now we can read as :  The preservation of both PT  and MPT, and TW values in the present cohort therefore supports the interpretation that these players could exhibit a robust and fatigue-resistant strength profile, characteristic of healthy elite soocer players.

 

Comment 44 : Page 10, Line 340–344: Limitations section is incomplete; include statistical and methodological limitations (e.g., multiple testing, lack of randomization).

 

Response 44 : Thank you for your comment, we have amended it accordingly and now we can read as : The cross-sectional design of the present study limits the ability to establish causal relationships between eccentric knee flexion isokinetic variables and performance out-comes or injury risk. While the observed velocity-dependent differences provide valuable descriptive and reliability information, longitudinal designs are required to determine whether these mechanical characteristics are predictive of future hamstring strain injuries or meaningful changes in on-field performance. In addition, although two commonly used eccentric angular velocities (60°·s⁻1 and 180°·s⁻1) were assessed, higher velocities (e.g., ≥300°·s⁻1), which may more closely replicate the rapid muscle–tendon loading experi-enced during maximal sprinting, were not included and could offer further insight into high-speed neuromuscular function. Finally, the absence of complementary neuro-muscular or structural assessments (such as electromyography, muscle architecture im-aging, or muscle–tendon stiffness measurements) limited the mechanistic interpretation of the velocity-related changes observed in PT, APT and TW, as the underlying neural and morphological contributors could not be directly examined. Additionally, the ob-served standard error of measurement (SEM) values were relatively low. While this likely reflects the standardized testing protocol, participant familiarization, and the use of a homogeneous sample of professional soccer players, it is also possible that methodological factors—such as the limited number of repetitions per condition—contributed to reduced variability. Therefore, the reported SEM values may represent both true measurement precision and potential methodological bias. Future studies should consider evaluating reliability across multiple variables and larger, more heterogeneous samples to better characterize the generalizability of these findings.

 

Comment 45 : Page 10, Line 346–348: The exclusion of higher velocities (≥300°·s⁻¹) is noted but not justified in design; clarify rationale.

 

Response 45 : Thank you for your comment. We acknowledge that higher angular velocities (≥300°·s⁻¹) were not included in the study. The rationale for this choice is twofold. First, velocities of 60°·s⁻¹ and 180°·s⁻¹ are the most commonly used in clinical and sports settings for isokinetic assessment of hamstring and quadriceps function, particularly in professional soccer players, and are well-supported in the literature for evaluating eccentric and concentric performance. Second, higher velocities are associated with greater technical challenges and increased risk of injury, especially during eccentric testing, and may not provide additional clinically meaningful information for this population. Therefore, our design prioritized safety, standardization, and clinical relevance.

 

Comment 46 : Page 10, Line 350–353: The absence of EMG or imaging limits interpretation; this should be emphasized earlier as a limitation.

 

Response 46 : Thank you for your comment. We included in the limitations of the study.

 

Comment 47 : Page 10, Line 355–360: Conclusions restate results but extend to injury prevention without evidence; reduce overinterpretation.

 

Response 47 : Thank you for your comment, we have amended it accordingly and now we can read as : Collectively, these findings indicate that a single eccentric isokinetic test is insufficient to fully characterize hamstring function in professional soccer players. Accordingly, the integration of multiple eccentric velocities and complementary variables such as PT, MPT, TW and angle-specific measures is recommended to enhance the sensitivity of isokinetic assessment in footballers.

 

Comment 48 : Page 10, Line 365–368: Recommendation of multiple velocities is reasonable but should be framed as preliminary.

 

Response 48 : Thank you for your comment.

 

Comment 49 : Page 10, Line 371–374: Practical applications suggest clinical relevance without validation; moderate strength of recommendations.

 

Response 49 : Thank you for your comment, we have amended it accordingly and now we can read as : The high inter-limb symmetry observed suggests that small side-to-side differences in peak torque or H:Q ratios may have clinical relevance in healthy professional soccer players. The present findings suggest that eccentric knee flexion strength testing at both 60°·s⁻1 and 180°·s⁻1 reflects distinct neuromuscular capacities. While 60°·s⁻1 provides in-sight into maximal force-generating capacity, higher angular velocities more closely reflect sprint-specific demands and may reveal mechanical behaviors (e.g., reduced peak torque or shifts in the angle of peak torque) that are not evident at slower speeds.

The excellent reliability and low SEM values indicate that eccentric knee flexion testing using five repetitions is suitable for tracking meaningful neuromuscular changes over time.

 

Comment 50 : Page 10, Line 385–387: MDC interpretation is appropriate but lacks context of longitudinal monitoring; clarify applicability.

 

Response 50 : Thank you. What is MDC ?

 

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the invitation to review the article "Velocity-dependent eccentric knee-flexion isokinetic assessment in elite professional soccer players: reliability, inter-limb asymmetry and mechanical characteristics". Please find detiled report below: 

1. You underlined in the abstract that some reliability metrics were calculated and used in abbreviated terms. I think those terms could not be familiar to everyone, so I propose two ways: 1) you could delete abbreviations and leave only the phrase that you calculated reliability metrics, or 2) explain those abbreviations.
2. Your results in the abstract are lacking the exact statistics. Any p-values or supporting data are welcome. 
3. As it is a cross-sectional study, I recommend attaching a strobe checklist. 
4. Did you analuzed power of your sample? Are you sure that 40 subjects will reach the required power? 
5. Which type of ICC (1, 2, or 3) did you use?
6. 6. If you present t-tests results in Table 1, the data should be supported with t statistic and degrees of freedom.
7. Do you have any additional data for your subjects except of age? There is a very little information provided and I'm quite sure that professional athletes are measdured more precisely.
8. Which type of athlete are your subjects according to McKay et al. system (doi: 10.1123/ijspp.2021-0451)?
9. Key paper from this domain was ommited and should be referenced in the discussion (doi: 10.3390/jcm11082077).

My recommendation is "Reconsider after major revision (substantial revisions to text or experimental methods needed)".

Author Response

Comment 1 : Thank you for the invitation to review the article "Velocity-dependent eccentric knee-flexion isokinetic assessment in elite professional soccer players: reliability, inter-limb asymmetry and mechanical characteristics". Please find detiled report below:

Response 1 : Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.

Comment 2 : You underlined in the abstract that some reliability metrics were calculated and used in abbreviated terms. I think those terms could not be familiar to everyone, so I propose two ways: 1) you could delete abbreviations and leave only the phrase that you calculated reliability metrics, or 2) explain those abbreviations.

Response 2 : Thank you for your comments. We do it the first option.

Comment 3 : Your results in the abstract are lacking the exact statistics. Any p-values or supporting data are welcome.

Response 3 : Thank you for your comment, we have amended it accordingly and now we can read as : Results: No significant inter-limb differences were observed for most variables (trivial–small effect sizes), except for higher eccentric TW at 180°·s⁻¹ in the dominant limb (p = 0.009). Eccentric PT and PTM decreased at higher velocities in both dominant (p = 0.002 and p<0.001, respectively) and non-dominant limbs (p = 0.008 and p<0.001, respectively), while APT shifted toward more flexed knee angles (p<0.001).

Comment 4 : As it is a cross-sectional study, I recommend attaching a strobe checklist.

Response 4 : Thank you for this suggestion. A completed STROBE checklist for cross-sectional studies has now been included as a supplementary file, ensuring that all relevant reporting items are adequately addressed in the manuscript.

Comment 5 : Did you analuzed power of your sample? Are you sure that 40 subjects will reach the required power?

Response 5 : Thanks to the reviewer for the recommendation. This is a specific, naturally occurring group measured throughout the 2023–2024, 2024–2025, and early 2025–2026 seasons, for which exclusion criteria were established for data analysis. Therefore, all players who did not meet the exclusion criteria were included. Even so, we have introduced the seasons in which they were evaluated within the text and it can be read like this: "Forty male professional soccer players (23.7 ± 3.5 years) were recruited from a Swiss first-division soccer club competing in the UEFA Conference League during 2023-2024, 2024-2025 and 2025-2026 seasons. The term “healthy professional soccer players” was used based on the exclusion criteria: Lack of full availability for regular participation in organized training and matches; a lower-limb injury in last 6 weeks; history of lower-limb surgery in the previous 12 months; current pain or symptoms preventing maximal testing and any medical conditions contraindicating high-intensity neuromuscular testing. »

Regarding power analysis,iIt is well documented by several articles that post hoc power calculations could not be useful (Althouse, 2020; Goodman & Berlin, 1994; Hoenig & Heisey, 2001). Based on this, and on the fact that we have tables within the results with a lot of information (change in mean, effect sizes (ES) and P value). We think that we will overload the table with more information, and it will not provide substantial information in addition to the sum of ES + P value. However, if the reviewer considers that we have to include it, please confirm it and we will do it. Thanks again.

Comment 6 : Which type of ICC (1, 2, or 3) did you use?

Response 6 : ICC (2,1)

 Comment 7 : If you present t-tests results in Table 1, the data should be supported with t statistic and degrees of freedom.

Response 7 : Thank you for this comment. We agree that reporting the t statistic and degrees of freedom can improve statistical transparency. However, given the number of variables and comparisons included in Table 1, adding these parameters would substantially increase the size and complexity of the table, potentially reducing its readability. To maintain clarity and accessibility of the results, we have chosen to report the most relevant information (mean ± SD, p-values, and effect sizes), which are commonly presented in similar studies. Nevertheless, detailed statistical outputs (including t values and degrees of freedom) can be provided upon request or included as supplementary material if required.

Comment 8 : Do you have any additional data for your subjects except of age? There is a very little information provided and I'm quite sure that professional athletes are measdured more precisely.

Response 8 : Thanks for the comment. Additional data have been included in the characteristics of the athletes “Forty male professional soccer players (23.7 ± 3.5 years; 78.5 ± 6.3 kg; 182.1 ± 7.5 cm; 47.9 ± 5.8 mm (sum of 6 skinfolds)) were recruited from a Swiss first-division soccer club competing in the UEFA Conference League during 2023-2024, 2024-2025 and 2025-2026 seasons”

 

Comment 9 : Which type of athlete are your subjects according to McKay et al. system (doi: 10.1123/ijspp.2021-0451)?

 

Response 9 : Thank you for your comment : 29 footballer were Tier 3: Highly Trained/National Level, and 11 were Tier 4: Elite/International Level;

 

Comment 10 Key paper from this domain was ommited and should be referenced in the discussion (doi: 10.3390/jcm11082077).

 

Response 10 : Thank you for your recomendation. This study examined the effectiveness of pressotherapy on diferents parameters that are beyond the scope of this study.

 

 

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Although the revised manuscript shows some improvements in clarity and presentation, the core scientific and methodological concerns remain insufficiently addressed, and these continue to undermine the validity and interpretability of the findings. Most critically, the statistical approach is still fundamentally flawed: the study relies on numerous independent comparisons across multiple dependent variables without any correction for multiple testing, while the authors explicitly acknowledge that no adjustment was applied. This substantially increases the risk of Type I error and calls into question the reliability of the few statistically significant findings reported. In addition, the justification regarding sample size and study planning remains inadequate, as the authors avoid addressing whether the design is sufficiently powered or appropriately structured to support their inferences. Several methodological responses are also conceptually weak or incomplete, including the lack of transparent reporting of assumption testing and the reliance on general practice rather than rigorous justification. Furthermore, the manuscript continues to overinterpret the results: despite being based on a cross-sectional, single-session design with no injury or longitudinal data, the discussion and practical applications still imply relevance for injury prevention, screening, and performance monitoring, which is not supported by the data and remains speculative. While some language has been softened, the overall narrative still extends beyond what the design can legitimately support. Reporting quality also remains below the expected standard, with inconsistent statistical detail, incomplete presentation of key metrics, and lingering issues in clarity and structure that limit reproducibility and critical appraisal. Finally, the study itself offers limited novelty, as it is largely descriptive and confirms well-established patterns without providing meaningful new mechanistic, analytical, or applied insight. Taken together, these unresolved issues indicate that the revision does not adequately address the major concerns raised in the original review.

Author Response

Comment 1 : Although the revised manuscript shows some improvements in clarity and presentation, the core scientific and methodological concerns remain insufficiently addressed, and these continue to undermine the validity and interpretability of the findings. Most critically, the statistical approach is still fundamentally flawed: the study relies on numerous independent comparisons across multiple dependent variables without any correction for multiple testing, while the authors explicitly acknowledge that no adjustment was applied. This substantially increases the risk of Type I error and calls into question the reliability of the few statistically significant findings reported. In addition, the justification regarding sample size and study planning remains inadequate, as the authors avoid addressing whether the design is sufficiently powered or appropriately structured to support their inferences. Several methodological responses are also conceptually weak or incomplete, including the lack of transparent reporting of assumption testing and the reliance on general practice rather than rigorous justification. Furthermore, the manuscript continues to overinterpret the results: despite being based on a cross-sectional, single-session design with no injury or longitudinal data, the discussion and practical applications still imply relevance for injury prevention, screening, and performance monitoring, which is not supported by the data and remains speculative. While some language has been softened, the overall narrative still extends beyond what the design can legitimately support. Reporting quality also remains below the expected standard, with inconsistent statistical detail, incomplete presentation of key metrics, and lingering issues in clarity and structure that limit reproducibility and critical appraisal. Finally, the study itself offers limited novelty, as it is largely descriptive and confirms well-established patterns without providing meaningful new mechanistic, analytical, or applied insight. Taken together, these unresolved issues indicate that the revision does not adequately address the major concerns raised in the original review.

 

Response 1 : Thank you very much for taking the time to review this manuscript. I will try to respond to each of the comments you make in your review.

 

Thank you for this important and thoughtful comment about statistics. We acknowledge the concern regarding the potential inflation of Type I error due to multiple comparisons. However, we would like to clarify that the statistical approach used in this study was hypothesis-driven and based on a predefined set of outcome variables commonly employed in isokinetic assessment (e.g., peak torque, mean peak torque, angle of peak torque, total work, and rate-of-torque-development–related variables), rather than on exploratory analyses. Importantly, these variables represent different but complementary dimensions of neuromuscular performance and are not independent outcomes in a strict statistical sense. Therefore, applying a global correction across all variables (e.g., Bonferroni) could be overly conservative and increase the risk of Type II error, potentially masking meaningful physiological differences. In addition, for analyses involving repeated measures (e.g., comparisons across repetitions), appropriate adjustments (Bonferroni post hoc tests and Greenhouse–Geisser correction when necessary) were applied to control for inflated error rates within each analytical framework.

 

Regarding the justification of sample size, we acknowledge that the justification of sample size and study planning required further clarification. The present study was conducted on a naturally occurring cohort of professional soccer players from a first-division team, where the sample size was determined by the availability of players meeting the inclusion criteria rather than by a priori recruitment targets. While no formal a priori power calculation was performed, the sample size (n = 40) is comparable to or larger than those reported in similar studies investigating isokinetic strength and neuromuscular performance in professional soccer players (e.g., Hamstring-to-Quadriceps Torque Ratios of Professional Male Soccer Players: A Systematic Review; Hamstring strain injuries: factors that lead to injury and re-injury). In addition, we have reported effect sizes alongside p-values ​​to provide a more comprehensive interpretation of the results, independent of sample size considerations. Based on the available sample (n = 40), an alpha level of 0.05, and assuming a small effect size (Cohen’s d = 0.3), the estimated statistical power was approximately 0.6. This indicates that the study is moderately powered to detect small effects, while being sufficiently powered for moderate to large effects. Given that many of the observed effects in this study were within the small-to-moderate range, the reporting of effect sizes alongside p-values provides important complementary information for interpretation.

Regarding the lack of transparent reporting on the testing of assumptions and the reliance on general practice rather than rigorous justification, we need you to be explicit about what you are referring to. We have attempted to respond to all the comments you made in the first review and have not identified which one you are addressing with this argument.

 

Concerning the ongoing overinterpretation of the results and the inclusion of the discussion on injuries, we have accepted all the corrections made in review 1. In fact, regarding the discussion on injury prevention, we have no intention of discussing injuries, as I mentioned in the first review, and therefore we are removing any comments that could be interpreted as a discussion of injuries:

 

In paragraph 4 of the Discussion, we stated: "Collectively, these findings reinforce the value of combining torque- and work-based variables to achieve a more comprehensive characterization of knee flexion–extension function, improving differentiation between uninjured and previously injured populations and enhancing the sensitivity of isokinetic assessments for performance monitoring and injury risk evaluation." and now we say "Collectively, these findings reinforce the value of combining torque- and work-based variables to achieve a more comprehensive characterization of knee flexion–extension function, improving the sensitivity of isokinetic assessments for performance monitoring".

 

 In paragraph 5 we have removed the following sentences which were based on previously published scientific studies: "Similar PT values ​​at 60°·s⁻1 have been reported in uninjured professional soccer players, whereas reduced eccentric strength at higher velocities has been associated with a history of hamstring injury and increased risk of reinjury [11, 16]."; "In rehabilitation contexts, deficits in eccentric total work (particularly at higher velocities) have been associated with incomplete recovery following hamstring injury [16, 20]."; "and hamstring injury mechanisms."

 

In paragraph 6 we have removed the following sentence (with which we do not 100% agree, since in many cases it may be that isokinetic assessments are used for this purpose in elite football; however, we have nevertheless removed it from the manuscript): "which is particularly relevant for monitoring training adaptations, rehabilitation progress, and return-to-play decision-making in elite soccer contexts".

Reviewer 2 Report

Comments and Suggestions for Authors

I do not have any further comments. 

Author Response

Comment 1 : I do not have any further comments. 

Response 1 : Thank you very much for taking the time to review this manuscript.