Skip to main content
NCBI home page
European Journal of Sport Science logoLink to European Journal of Sport Science
. 2026 Oct 5;26(11):e70264. doi: 10.1002/ejsc.70264

Acute Reactive Strength Responses and Peak Power–Performance Relationships After Flywheel Calf Raises in National‐Level Track‐and‐Field Athletes: A Randomized Crossover Trial

Chen Chen 1,2, Tian Huang 3, Junlong Zhang 4, Bing Yan 5,✉, Olivier Girard 6
PMCID: PMC13635872  PMID: 42831264

ABSTRACT

This study examined reactive strength responses to flywheel eccentric calf raises performed under three inertial loads and the associations between peak power and post‐intervention reactive strength. Thirty‐seven national‐level male track‐and‐field athletes completed three randomized crossover trials (≥ 48 h apart), each involving 3 × 8 calf raises under Light (0.0281 kg·m2), Moderate (0.0711 kg·m2), or High (0.1011 kg·m2) inertia. Reactive strength was assessed pre‐ and 4‐min post‐exercise using a 10/5 repeated‐jump test measuring ground contact time (GCT), jump height (JH), and reactive strength index (RSI). Concentric and eccentric peak power (ConPP, EccPP) were recorded. JH and RSI increased after Light (p = 0.018 and p < 0.001) and Moderate (both p < 0.001), while GCT decreased after both conditions (both p < 0.001). No statistically clear improvement was detected after High. Moderate produced greater RSI improvements than Light (p = 0.008) and High (p < 0.001), and greater JH gains than Light (p = 0.020) and High (p = 0.016). Light produced greater RSI gains than High (p = 0.022), while GCT decreased more after Light and Moderate than High (both p < 0.001). Power–performance associations varied across loads, with significant ConPP × Load interactions for JH (p = 0.002) and RSI (p = 0.003), and EccPP × Load interactions for JH (p = 0.040) and RSI (p = 0.025). Under High load, ConPP was associated with JH (β = 1.37 cm, 95% CI 0.57–2.16) and RSI (β = 0.086, 95% CI 0.020–0.153), while EccPP was associated with JH (β = 1.27 cm, 95% CI 0.42–2.12; all p ≤ 0.012). Overall, greater improvements were observed 4‐min after light and particularly moderate inertia, whereas high‐load peak power was more strongly related to post‐intervention performance.

Keywords: ankle plantar flexor, flywheel training, peak power output, post‐activation performance enhancement, reactive strength index

Highlights

  • Reactive strength improved at 4 min after Light and Moderate flywheel calf raises in national‐level male track‐and‐field athletes, whereas no statistically clear improvement was detected after High inertia.

  • Moderate inertia produced the most consistent improvements in jump height and reactive strength index 4‐min post‐intervention.

  • Peak power–performance relationships were load dependent, with stronger associations between concentric and eccentric peak power and post‐intervention reactive strength observed under High inertia.

1. Introduction

Track‐and‐field performance relies heavily on reactive strength (Southey et al. 2024). Reactive strength reflects the ability to rapidly absorb and reapply force during explosive stretch‐shortening cycle (SSC) actions (Beattie et al. 2017). Practically, it denotes the capacity to produce large propulsive forces within a brief force‐application window, minimize ground contact time (GCT), or both (Douglas et al. 2018). Accordingly, jumping and sprinting performance depend on generating large propulsive forces while minimizing contact duration to maximize mechanical output and movement efficiency (Nagahara and Zushi 2016; Nagahara et al. 2018; Weyand et al. 2000). The reactive strength index (RSI) is a widely used field‐based indicator of reactive strength. Typically derived from drop‐jump performance, RSI is calculated by dividing jump height (JH) by GCT (Comfort et al. 2019; Ebben and Petushek 2010). Higher RSI values are associated with superior sprinting ability and greater power output during loaded jumps, supporting its utility for performance monitoring and athlete profiling (Jarvis et al. 2022; Kipp et al. 2017; Southey et al. 2024).

Reactive strength can be enhanced through long‐term training or acutely through post‐activation performance enhancement (PAPE). PAPE is typically induced by brief, high‐load conditioning activities performed before explosive tasks and reflects transient improvements in neuromuscular function (Blazevich and Babault 2019; Seitz and Haff 2016). High‐intensity eccentric loading appears particularly effective for eliciting PAPE, potentially through increased motor unit recruitment, greater musculoskeletal stiffness, and improved force transmission (Blazevich and Babault 2019; Seitz and Haff 2016). In this context, inertial flywheel devices have gained attention as a practical means of imposing substantial eccentric demands, particularly during the braking phase of each repetition (Beato and Dello Iacono 2020; De Keijzer et al. 2020). Although flywheel conditioning can acutely enhance jumping and sprinting performance, most existing evidence is derived from squat‐based protocols, limiting insight into joint‐specific applications (Beato et al. 2021; McErlain‐Naylor and Beato 2021; Wren et al. 2023).

Given the importance of ankle function during sprinting and rapid SSC actions, the distal musculotendinous system represents a key target for acute potentiation (Nagahara et al. 2018). Calf raises specifically target the ankle plantar flexors, which contribute substantially to propulsion and elastic energy return during SSC actions (Crotty et al. 2024; Lai et al. 2016). Shorter contact times are associated with higher plantar‐flexor force capacity and Achilles tendon stiffness (Abdelsattar et al. 2018), providing a rationale for ankle‐focused conditioning to acutely enhance RSI (Beato and Dello Iacono 2020). However, it remains unclear whether an ankle‐dominant flywheel task, such as the flywheel calf raise, can elicit meaningful PAPE responses in highly trained track‐and‐field athletes.

Flywheel devices provide real‐time measures of concentric and eccentric peak power outputs, which are commonly used to monitor training load and fatigue (Maroto‐Izquierdo et al. 2021; Wren et al. 2023). Peak power is sensitive to inertial load, with higher inertia generally reducing peak power and altering concentric–eccentric output characteristics (Martinez‐Aranda and Fernandez‐Gonzalo 2017; Piqueras‐Sanchiz et al. 2020). Consequently, different inertial loads may produce distinct acute performance responses and modify the relationship between mechanical output during the conditioning activity and subsequent explosive performance. Despite their widespread use, the ecological validity of these metrics as predictors of sport‐specific outcomes (i.e., jumping and sprinting) remains unclear (Maroto‐Izquierdo et al. 2021). Finally, PAPE research using flywheel devices is largely dominated by squat‐derived protocols rather than ankle‐dominant tasks (Beato and Dello Iacono 2020; McErlain‐Naylor and Beato 2021). As such, it remains unknown whether peak power produced during flywheel calf raises relates to subsequent reactive strength and whether these associations vary across inertial loads.

Therefore, this study aimed to determine whether a flywheel calf‐raise protocol acutely enhances reactive strength in track‐and‐field athletes, and whether concentric and eccentric peak power outputs across three inertial loads are associated with post‐intervention reactive strength. We hypothesized that (i) reactive‐strength responses would be load‐dependent, with moderate and high inertial loads producing greater improvements than Light load; and (ii) both concentric and eccentric peak power would be positively associated with reactive strength, with the strength of these associations varying across inertial loads.

2. Methods

2.1. Sample Size Calculation

An a priori power calculation was performed using G*Power (v3.1.9.6; University of Düsseldorf, Düsseldorf, Germany) for the primary Load × Time interaction in reactive strength. Assuming a medium effect (f = 0.25), α = 0.05, power (1−β) = 0.80, a repeated‐measures correlation of 0.50, and a nonsphericity correction of ε = 0.75, the minimum required sample was 34 participants. To account for potential attrition, 37 participants were recruited.

2.2. Participants

Thirty‐seven male national‐level track‐and‐field athletes (20 triple jumpers and 17 long jumpers) participated. They had at least 2 years of resistance training experience (mean ± SD, age: 19.9 ± 3.1 years; body weight: 71.1 ± 4.0 kg; height: 182.2 ± 5.0 cm; resistance training experience: 3.4 ± 2.0 years). Based on their competition level, participants were classified as Tier 3 (Highly Trained/National Level) according to McKay et al. (2022). All were healthy, free from lower‐limb injury, and familiar with flywheel devices through regular training. The study protocol was approved by the institutional ethics committee (XDYX202604K15), and written informed consent was obtained from all participants. All procedures comply with the Declaration of Helsinki.

2.3. Study Design

A randomized, crossover design was used to investigate the acute effects of different inertial loads during flywheel calf raises on reactive strength. Participants first completed a familiarization session to practice both the flywheel calf‐raise intervention and reactive strength assessments. They then attended the laboratory on three separate occasions over a 2‐week period. All sessions were conducted at the same time of day (± 1 hour) and separated by at least 48 hours.

Each session followed a standardized sequence: (1) warm‐up consisting of 5‐min self‐paced jogging followed by dynamic stretching; (2) pre‐intervention reactive strength assessment; (3) flywheel calf‐raise eccentric intervention executed at one of three randomized inertial loads; and (4) post‐intervention reactive strength assessment conducted 4 minutes after the intervention. The 4‐min recovery interval was selected a priori to capture the early PAPE window reported following flywheel conditioning (Fu et al. 2023; Shi et al. 2024). Previous work suggests that > 3 min of recovery is typically required to elicit meaningful PAPE responses (De Keijzer et al. 2020). A single post‐test time point was used to minimize fatigue from repeated jumping assessments. The load order was randomized within participants to reduce learning and order effects. All sessions were conducted in the same indoor venue, by the same investigators, and under identical conditions. Participants wore identical clothing and footwear across sessions.

2.4. Procedures

Reactive strength assessment: Reactive strength was assessed using the 10/5 repeated‐jump test (Stratford et al. 2021) performed on a contact mat (Smartjump, Fusion Sport, Coopers Plains, Australia). Participants executed one countermovement jump followed immediately by 10 consecutive maximal bilateral hops with hands on hips to standardize technique. Standardized verbal cues were provided to “jump as high as possible” and “minimize ground‐contact time”. A maximum GCT threshold of < 250 ms was applied to identify fast‐SSC contacts (Stratford et al. 2021). The five contacts with the highest RSI values were identified, and GCT, JH and RSI were averaged across these five same contacts for statistical analysis.

Flywheel calf‐raise training: Participants performed bilateral calf raises on a flywheel device with interchangeable inertia (D11 full; Desmotec, Biella, Italy). Three inertial loads were tested on separate visits: 0.0281 kg m2 (Light), 0.0711 kg m2 (Moderate), and 0.1011 kg m2 (High). Each intervention comprised three sets of eight maximal repetitions with 3 min of passive rest between sets (Shi et al. 2024). Both feet were positioned centrally on the platform, the flywheel cord was attached to a waist belt, and participants grasped the front horizontal bar with both hands for balance. Participants performed the concentric phase at maximal intended velocity, moving from a flat‐foot position to maximal plantar flexion, and then actively resisted the eccentric return to the starting position. Slight knee flexion (10°–20°) was maintained to avoid joint locking and was visually monitored by the same investigator. Real‐time visual feedback was provided via the device display throughout the exercise. For each participant and inertial condition, the highest concentric and eccentric peak power outputs (ConPP and EccPP, respectively) recorded across the 24 maximal repetitions were retained for analysis. The two submaximal acceleration repetitions performed before each set to accelerate the flywheel were excluded, but no maximal repetitions were removed.

2.5. Statistical Analyses

Analyses were performed using SPSS (version 31.0; IBM Corp., Armonk, NY, USA). Data are presented as mean ± standard deviation (SD). Normality was evaluated via the Shapiro–Wilk test. Acute effects of flywheel conditioning on reactive strength were analyzed using a 3 (load: Light, Moderate, and High) × 2 (time: pre‐ and post‐tests) repeated‐measures ANOVA. Individual percentage changes in each reactive strength outcome and eccentric overload index were calculated for each participant and load as [(Post − Pre)/Pre] × 100 and [(EccPP/ConPP) − 1] × 100, respectively. Each variable was compared across loads using a separate one‐way repeated‐measures ANOVA. Sphericity was evaluated using Mauchly's test, with Greenhouse–Geisser‐adjusted degrees of freedom and p values reported when the sphericity assumption was violated (Table S1). When significant main effects or interactions were observed, Holm‐adjusted pairwise comparisons were performed, including within‐load (pre vs. post) and between‐load (change‐score) contrasts. Effect sizes for ANOVA effects were reported as partial eta squared (ηp2), and pairwise comparisons as Cohen's dz.

Associations between peak power output and reactive strength outcomes were examined using linear mixed‐effects models (LMM), with participant specified as a random intercept. Post‐intervention values (GCT, JH, and RSI) were entered as dependent variables, with corresponding pre‐intervention values as covariates and load (reference category = Light) as a fixed factor. Due to high collinearity between EccPP and ConPP (r = 0.886), separate models were constructed. Load‐dependent associations were tested by adding a Power × Load interaction, with interaction and main‐effects models compared via likelihood‐ratio tests (df = 2) (Table 1, Panel B). Models used for likelihood‐ratio comparisons were fitted using maximum‐likelihood estimation. Load‐stratified models were then used to estimate within‐load slopes and 95% CIs. To facilitate interpretation, slopes represent the expected difference in the post‐intervention outcome associated with a 1‐SD increase in peak power within each load condition, adjusted for the corresponding pre‐intervention value (Figure 1). Full fixed‐effect estimates, variance components, and model diagnostics are provided in Table S2. Statistical significance was set at p < 0.05.

TABLE 1.

Peak power outputs across loads and likelihood‐ratio tests for load‐dependent power‐reactive strength associations.

A. Peak power output and eccentric overload B. LRT p values (power × load interaction)
Variable Light Moderate High GCT (p) JH (p) RSI (p)
EccPP (kW) 1.520 ± 0.306 1.853 ± 0.347 1.774 ± 0.495 0.604 0.040* 0.025*
ConPP (kW) 1.444 ± 0.298 1.663 ± 0.283 1.605 ± 0.317 0.439 0.002** 0.003**
Eccentric overload (%) 6.08 ± 11.95 11.57 ± 11.02 9.10 ± 13.49
Open in a new tab

Note: Panel A (left) presents peak power outputs (mean ± SD) recorded during the flywheel calf‐raise under each load condition (n = 37). Eccentric overload was calculated separately for each participant and condition as [(EccPP/ConPP) − 1] × 100. Panel B (right) reports p values from likelihood‐ratio tests (LRT; df = 2) comparing nested linear mixed‐effects models (maximum likelihood estimation) to evaluate whether the association between peak power and reactive strength outcomes differed across loads (Power × Load interaction). For each outcome (ground contact time [GCT], jump height [JH], and reactive strength index [RSI]), the post‐intervention value was modeled with the corresponding pre‐intervention value as a covariate, with participant included as a random intercept.

Abbreviations: ConPP, concentric peak power (kW); EccPP, eccentric peak power.

*

p < 0.05.

**

p < 0.01.

FIGURE 1.

FIGURE 1

Open in a new tab

Load‐specific associations between eccentric/concentric peak power and reactive strength outcomes. Points represent load‐specific slopes (β per 1‐SD increase in peak power within each load) derived from load‐stratified linear mixed‐effects models and adjusted for the corresponding pre‐intervention value. Horizontal bars indicate 95% confidence intervals. Asterisks denote slopes that differ significantly from zero (i.e., indicating a significant power–outcome association at that load). *p < 0.05, **p < 0.01.

3. Results

3.1. Acute Effects on Reactive Strength

Pre‐ and post‐intervention descriptive statistics and mean changes with 95% CIs are presented in Table 2. Baseline values did not differ across the three sessions for GCT (p = 0.705), JH (p = 0.724), or RSI (p = 0.379).

TABLE 2.

Descriptive statistics and within‐condition changes in reactive‐strength outcomes.

Outcome Load Pre, mean ± SD Post, mean ± SD Mean change ± SD [95% CI] Holm‐adjusted p Cohen's dz
GCT (ms) Light 170.58 ± 15.28 161.24 ± 12.67 −9.34 ± 7.58 [−11.87, −6.82] < 0.001 −1.23
Moderate 170.04 ± 13.83 159.65 ± 12.10 −10.39 ± 5.12 [−12.10, −8.69] < 0.001 −2.03
High 170.94 ± 9.94 168.96 ± 13.87 −1.97 ± 7.92 [−4.61, 0.67] 0.138 −0.25
JH (cm) Light 36.19 ± 3.29 36.99 ± 4.16 0.80 ± 1.76 [0.21, 1.39] 0.018 0.45
Moderate 36.08 ± 3.38 37.63 ± 3.83 1.55 ± 1.40 [1.08, 2.02] < 0.001 1.11
High 36.17 ± 3.06 36.55 ± 4.61 0.38 ± 2.79 [−0.55, 1.31] 0.412 0.14
RSI Light 2.154 ± 0.267 2.325 ± 0.368 0.171 ± 0.167 [0.116, 0.227] < 0.001 1.03
Moderate 2.142 ± 0.285 2.378 ± 0.347 0.236 ± 0.086 [0.207, 0.264] < 0.001 2.76
High 2.129 ± 0.222 2.204 ± 0.420 0.076 ± 0.253 [−0.009, 0.160] 0.078 0.30
Open in a new tab

Note: Values are mean ± SD (n = 37). Mean change was calculated as Post − Pre. p values are Holm‐adjusted for within‐condition pre–post comparisons. Cohen's dz was calculated using paired change scores.

Abbreviations: CI, confidence interval; GCT, ground contact time; JH, jump height; RSI, reactive strength index.

Significant load × time interactions were observed for GCT (p < 0.001, ηp2 = 0.328), JH (p = 0.013, ηp2 = 0.124), and RSI (p < 0.001, ηp2 = 0.245). GCT decreased after Light (Δ = −9.34 ms, p < 0.001, dz = −1.23) and Moderate (Δ = −10.39 ms, p < 0.001, dz = −2.03), while no statistically clear change was detected after High (p = 0.138). JH increased after Light (Δ = 0.80 cm, p = 0.018, dz = 0.45) and Moderate (Δ = 1.55 cm, p < 0.001, dz = 1.11), while no statistically clear change was detected after High (p = 0.412) (Figure 2A–C). Likewise, RSI increased after Light (Δ = 0.171, p < 0.001, dz = 1.03) and Moderate (Δ = 0.236, p < 0.001, dz = 2.76), while no statistically clear change was detected after High (p = 0.078).

FIGURE 2.

FIGURE 2

Open in a new tab

Acute changes in reactive strength after flywheel eccentric calf‐raise under different inertial loads (n = 37). (Panels A–C) show absolute pre–post change scores (Δ = Post − Pre) for ground contact time (A; ΔGCT, ms), jump height (B; ΔJH, cm), and reactive strength index (C; ΔRSI) across Light, Moderate, and High loads. (Panel D) shows the corresponding percentage changes for each outcome (Δ% = (Post − Pre)/Pre × 100). Scatter points represent individual participants, while diamonds and error bars indicate the mean change ± 95% confidence interval (CI). Asterisks above each load denote significant within‐load pre–post effects, and bracketed asterisks indicate significant between‐load differences in change scores (e.g., Light vs. High, Moderate vs. High, and/or Light vs. Moderate) for the connected conditions. Negative values indicate reductions (e.g., shorter GCT). *p < 0.05, **p < 0.01.

Post hoc comparisons revealed significant differences in change scores among loads (Figure 2A–C). Decreases in GCT were larger after Light than High (Δ = −7.37 ms, p < 0.001, dz = −0.75) and after Moderate than High (Δ = −8.42 ms, p < 0.001, dz = −0.84), with no difference between Light and Moderate (p = 0.449). Increases in JH were greater after Moderate than Light (Δ = 0.75 cm, p = 0.020, dz = 0.45) and High (Δ = 1.17 cm, p = 0.016, dz = 0.49), whereas Light and High did not differ (p = 0.332). Increases in RSI were greater after Moderate than Light (Δ = 0.065, p = 0.008, dz = 0.50) and High (Δ = 0.160, p < 0.001, dz = 0.73), and after Light than High (Δ = 0.096, p = 0.022, dz = 0.40).

Percentage changes showed a similar pattern. Under Light, Moderate, and High inertia, GCT decreased by 5.29%, 6.03%, and 1.21%; the reductions were greater after Light and Moderate than after High (both p < 0.001). JH increased by 2.10%, 4.29%, and 0.93%, respectively, with a greater increase after Moderate than after Light (p = 0.020) and High (p = 0.013). RSI increased by 7.68%, 10.88%, and 2.95% under Light, Moderate, and High loads, respectively; the increase was greater after Moderate than after Light (p = 0.005) and High (p < 0.001), and after Light than after High (p = 0.013). These findings were consistent with the absolute change‐score analyses (Figure 2D).

3.2. Power Output‐Performance Associations

Eccentric and concentric peak power outputs across loads are presented in Table 1 (Panel A), together with participant‐level eccentric‐overload values. Eccentric overload was 6.08 ± 11.95%, 11.57 ± 11.02%, and 9.10 ± 13.49% for the Light, Moderate, and High conditions, respectively, with no statistically significant difference among loads (p = 0.061). Power × load interaction terms from the linear mixed‐effects models are reported in Table 1 (Panel B). Significant interactions were observed for ConPP with JH (p < 0.01) and RSI (p < 0.01), indicating that the relationship between ConPP and these outcomes differed across loads. No interaction was found for ConPP with GCT (p = 0.439). For EccPP, significant interactions were observed with JH (p = 0.040) and RSI (p = 0.025), suggesting that the relationship between EccPP and these outcomes differed across loads. No significant interactions were found for EccPP with GCT (p = 0.604).

In load‐stratified models, neither ConPP nor EccPP was significantly associated with GCT under any load (all p ≥ 0.139; Figure 1A). Under High inertia, a 1‐SD increase in ConPP (0.317 kW) was associated with a 1.37‐cm higher post‐intervention JH (95% CI: 0.57–2.16; p = 0.001), with no significant associations under Moderate (p = 0.098) or Light (p = 0.787) (Figure 1B). A 1‐SD increase in ConPP under High inertia was also associated with 0.086‐unit higher post‐intervention RSI (95% CI: 0.020–0.153; p = 0.012), whereas no significant associations were observed under Moderate (p = 0.106) or Light (p = 0.851) conditions (Figure 1C).

For EccPP, a 1‐SD increase under High inertia (0.495 kW) was associated with a 1.27‐cm higher post‐intervention JH (95% CI: 0.42–2.12; p = 0.004), with no significant associations under Light (p = 0.460) or Moderate (p = 0.266) (Figure 1B). EccPP was not significantly associated with post‐intervention RSI under any condition (High: p = 0.083; Light/Moderate: p ≥ 0.321), despite the significant overall EccPP × Load interaction (Figure 1C).

4. Discussion

The main finding was that GCT decreased and JH and RSI increased 4 min after the Light and Moderate conditions, whereas no significant pre‐post changes were observed under High inertia. The most consistent improvements occurred after Moderate inertia. Therefore, our first hypothesis was only partly supported: reactive strength responses differed across inertial loads, with the largest improvements observed after Light and particularly moderate inertia rather than moderate inertia as hypothesized.

These load‐specific acute responses partly align with evidence that flywheel eccentric‐overload exercise can induce PAPE (Beato et al. 2020) but also indicate that the inertia associated with the clearest response may depend on the specific conditioning task and performance outcome. In flywheel half‐squat protocols, improvements in countermovement jump performance are typically observed under moderate and high inertia, with limited effects under light loads (Fu et al. 2023). Likewise, moderate‐to‐high inertia during flywheel exercise has been shown to enhance power‐related outcomes such as countermovement jump, long jump, and change‐of‐direction performance (Beato, De Keijzer, et al. 2021). In contrast, reactive strength improved after Light inertia in our study, whereas no statistically significant improvement was detected under High load at 4 min. A likely explanation is the specificity of the conditioning task. Previous studies primarily used multi‐joint squat or half‐squat exercises that emphasize hip and knee extensor output and involve greater global force production (Martinez‐Aranda and Fernandez‐Gonzalo 2017; Sabido et al. 2018). Here, the ankle‐dominant task and repeated‐jump outcome emphasized short‐contact SSC execution and plantar‐flexor muscle–tendon function (Abdelsattar et al. 2018; Crotty et al. 2024; Kipp et al. 2017; Lai et al. 2016). Under such conditions, lower inertial loads may provide a sufficient conditioning stimulus to improve reactive strength, although differences in total mechanical exposure across loads may also have contributed to the observed responses.

The absence of improvement under the High‐load condition suggests that heavier inertia may not be optimal at the 4‐min post‐test interval, although the underpinning mechanism cannot be determined from a single time point. This interpretation is consistent with the potentiation–fatigue framework, whereby performance benefits depend on whether potentiation exceeds residual fatigue at the time of reassessment (Seitz and Haff 2016). PAPE responses to flywheel protocols are also influenced by the interaction between intensity, volume, and recovery duration (Blazevich and Babault 2019; Seitz and Haff 2016). For instance, a near‐maximal flywheel squat protocol (1–3 sets × 6 reps) did not improve jump height 1–12 min post‐exercise, likely due to excessive fatigue (Hall et al. 2024). Eccentric‐overload half‐squat protocols have been shown to improve countermovement and long jump performance at 6 min, but not 3 min post‐exercise, suggesting that longer recovery may be required after multi‐set protocols (2–3 × 6 reps) (De Keijzer et al. 2020). However, another half‐squat study reported that both moderate (0.057 kg·m2) and high loads (0.122 kg·m2) improved countermovement performance 4–8 min post‐exercise (Fu et al. 2023). Similarly, flywheel squats performed with 3 × 8 reps at moderate (0.0784 kg·m2) and high loads (0.1568 kg·m2) improved countermovement performance immediately (0 min), with effects maintained 4–12 min post‐exercise (Shi et al. 2024). Importantly, the present conditions were matched for repetitions but not for total mechanical exposure. Similar CMJ and RSI improvements have been reported after low‐ and high‐inertia flywheel conditioning when total force impulse was matched, suggesting that the less apparent response after High inertia may partly reflect greater accumulated mechanical exposure or fatigue rather than inertia alone (Tsoukos et al. 2025). Taken together, these observations indicate that Light and Moderate conditions were associated with clearer improvements in reactive strength than the High condition in 4 min. However, longer recovery periods or total mechanical work matching are required to determine whether the absence of improvement under High load reflects residual fatigue or a delayed PAPE response.

Another novel finding was that power–performance associations were load‐dependent, with significant Power × Load interactions for ConPP with JH and RSI, and for EccPP with these outcomes. However, significant within‐load associations were observed only under High load, where higher ConPP was associated with greater post‐test JH and RSI, and higher EccPP was associated with greater post‐test JH. Accordingly, our second hypothesis was partially supported, suggesting that High inertia may better reveal meaningful power–performance relationships.

These interactions indicate that peak power is not a universal predictor of reactive strength across loads, but became more informative as the mechanical demands of the flywheel task increased with inertia. This aligns with evidence that flywheel inertia alters concentric and eccentric power outputs and their ratio, meaning that peak power reflects different mechanical constraints depending on load (de Keijzer et al. 2022). Lower inertia permits higher peak power expression, whereas higher inertia tend to increase eccentric braking demands, fundamentally altering neuromuscular requirements rather than simply scaling power output (O'Brien et al. 2022). More broadly, rotary‐inertial and traditional resistance systems can exhibit distinct force–velocity–displacement profiles even when concentric intensity is matched, reinforcing that mechanical outputs must be interpreted within the specific resistance configuration in which they are produced (Galiano et al. 2025).

In our study, the strongest associations emerged under High load, where higher peak power was related to higher post‐test JH (ConPP and EccPP) and RSI (ConPP). Notably, post‐intervention reactive strength outcomes also showed greater between‐athlete variability under this condition, with some participants improving substantially, while others showed minimal change or even decrements. This wider dispersion suggests that higher inertia can magnify individual differences in physical capability (Asencio et al. 2024; Spudić et al. 2024). Athletes able to sustain higher power under heavier inertia likely possess greater strength and neuromuscular capacity, both of which underpin reactive strength and SSC performance (Beattie et al. 2017; Douglas et al. 2018; Jarvis et al. 2022). Consistent with this interpretation, superior jump performance has been associated with greater concentric peak power during countermovement jumps (Barker et al. 2018), while greater eccentric force capacity is also linked to enhanced vertical jump performance (Nishiumi et al. 2023). Although the EccPP × Load interaction was significant for RSI, none of the within‐load EccPP–RSI slopes reached statistical significance, indicating that this interaction should be interpreted cautiously. Together, these observations provide a mechanistic rationale for the stronger ConPP associations with JH and RSI, and the stronger EccPP association with JH under high load. In contrast, GCT showed no meaningful association with peak power at any load. This is not unexpected, as contact time during fast SSC actions is largely influenced by stiffness regulation and tendon‐muscle mechanics rather than peak power alone (Hobara et al. 2007; Kipp et al. 2017). For instance, Achilles tendon stiffness is closely linked to contact time during jumping tasks (Abdelsattar et al. 2018). Therefore, peak power generated during the flywheel exercise may not directly translate to reductions in GCT, which likely depend more on stiffness regulation than on maximal power production (Souhaiel and Denis 2001).

Several limitations should be acknowledged. First, the absence of a non‐exercise control condition limits causal attribution of the observed pre–post changes because repeated testing and other time‐related effects cannot be excluded. Second, the three inertial conditions were matched for the number of repetitions rather than total mechanical exposure. Mechanical work, force impulse, time under tension, and repetition‐to‐repetition power loss were not measured. Consequently, differences in accumulated fatigue may have contributed to the load‐specific responses. Third, reactive strength was assessed only 4 min post‐conditioning; therefore, the study cannot determine whether the lack of a statistically clear improvement after High inertia reflected a genuinely absent response or potentiation emerging later. Fourth, mechanistic measures (e.g., surface EMG, tendon stiffness) were not collected, thereby interpretations of underlying physiological mechanisms remain speculative. Finally, the sample consisted exclusively of male national‐level long‐ and triple‐jump athletes, limiting generalizability to female athletes, other track‐and‐field disciplines, and other sports. Although flywheel conditioning has shown acute performance benefits in rugby union, ice hockey, and soccer (Murton et al. 2023; Puustinen et al. 2023; Raya‐González et al. 2021), it remains unclear whether the associations between peak power and reactive strength observed during flywheel calf raises extend to these populations.

5. Conclusion

Reactive‐strength responses observed 4 min after flywheel eccentric calf raises differed across inertial conditions. GCT decreased and JH and RSI increased after Light and Moderate inertia, whereas no statistically clear changes were observed after high inertial conditions; the most consistent improvements occurred after moderate inertial conditions. Peak power–performance relationships were also load‐specific. Under High load, greater ConPP was associated with higher post‐test JH and RSI, and greater EccPP was associated with higher JH, whereas neither metric was related to GCT. These findings suggest that both acute reactive‐strength responses and the associations between peak power and subsequent performance depend on the inertial load, although conclusions are restricted to the specific protocol and 4‐min assessment used.

Author Contributions

Chen Chen: data curation, formal analysis, methodology, software, validation, writing – original draft, writing – review and editing. Tian Huang: data curation, formal analysis, software, validation, writing – review and editing. Junlong Zhang: software, investigation, writing – original draft. Bing Yan: conceptualization, funding acquisition, project administration, resources, supervision, validation, writing – review and editing. Olivier Girard: validation, writing – review and editing. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the Fujian Provincial Social Science Fund (Grant FJ2026C100) and the Fundamental Research Funds for the Central Universities, China (Grant 2072021150).

Ethics Statement

The study protocol was approved by the Xiamen university ethics committee (XDYX202604K15), and written informed consent was obtained from all participants.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Sphericity assessment and repeated‐measures ANOVA results.

EJSC-26-e70264-s002.docx (28.6KB, docx)

Table S2: Linear mixed‐effects model estimates and diagnostics.

EJSC-26-e70264-s001.docx (36.5KB, docx)

Acknowledgements

The authors would like to express their sincere gratitude to all the athletes who dedicated their time and effort to participate in this study.

Data Availability Statement

The data that support the findings of this study are not publicly available owing to privacy and ethical restrictions concerning the athlete participants; however, they can be requested from the corresponding author upon reasonable request.

References

  1. Abdelsattar, M. , Konrad A., and Tilp M.. 2018. “Relationship Between Achilles Tendon Stiffness and Ground Contact Time During Drop Jumps.” Journal of Sports Science and Medicine 17, no. 2: 223–228. [PMC free article] [PubMed] [Google Scholar]
  2. Asencio, P. , García‐Valverde A., Albaladejo‐García C., Beato M., Moreno F. J., and Sabido R.. 2024. “Analysis of Concentric and Eccentric Power in Flywheel Exercises Depending on the Subjects' Strength Level and Body Mass.” Journal of Strength & Conditioning Research 38, no. 8: 1394–1400. 10.1519/JSC.0000000000004818. [DOI] [PubMed] [Google Scholar]
  3. Barker, L. A. , Harry J. R., and Mercer J. A.. 2018. “Relationships Between Countermovement Jump Ground Reaction Forces and Jump Height, Reactive Strength Index, and Jump Time.” Journal of Strength & Conditioning Research 32, no. 1: 248–254. 10.1519/JSC.0000000000002160. [DOI] [PubMed] [Google Scholar]
  4. Beato, M. , De Keijzer K. L., Leskauskas Z., Allen W. J., Dello Iacono A., and McErlain‐Naylor S. A.. 2021. “Effect of Postactivation Potentiation After Medium vs. High Inertia Eccentric Overload Exercise on Standing Long Jump, Countermovement Jump, and Change of Direction Performance.” Journal of Strength & Conditioning Research 35, no. 9: 2616–2621. 10.1519/JSC.0000000000003214. [DOI] [PubMed] [Google Scholar]
  5. Beato, M. , and Dello Iacono A.. 2020. “Implementing Flywheel (Isoinertial) Exercise in Strength Training: Current Evidence, Practical Recommendations, and Future Directions.” Frontiers in Physiology 11: 569. 10.3389/fphys.2020.00569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Beato, M. , McErlain‐Naylor S. A., Halperin I., and Dello Iacono A.. 2020. “Current Evidence and Practical Applications of Flywheel Eccentric Overload Exercises as Postactivation Potentiation Protocols: A Brief Review.” International Journal of Sports Physiology and Performance 15, no. 2: 154–161. 10.1123/ijspp.2019-0476. [DOI] [PubMed] [Google Scholar]
  7. Beato, M. , Stiff A., and Coratella G.. 2021. “Effects of Postactivation Potentiation After an Eccentric Overload Bout on Countermovement Jump and Lower‐Limb Muscle Strength.” Journal of Strength & Conditioning Research 35, no. 7: 1825–1832. 10.1519/JSC.0000000000003005. [DOI] [PubMed] [Google Scholar]
  8. Beattie, K. , Carson B. P., Lyons M., and Kenny I. C.. 2017. “The Relationship Between Maximal Strength and Reactive Strength.” International Journal of Sports Physiology and Performance 12, no. 4: 548–553. 10.1123/ijspp.2016-0216. [DOI] [PubMed] [Google Scholar]
  9. Blazevich, A. J. , and Babault N.. 2019. “Post‐Activation Potentiation Versus Post‐Activation Performance Enhancement in Humans: Historical Perspective, Underlying Mechanisms, and Current Issues.” Frontiers in Physiology 10: 1359. 10.3389/fphys.2019.01359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Comfort, P. , Jones P., and McMahon J. J.. 2019. Performance Assessment in Strength and Conditioning. Routledge, Taylor & Francis Group. [Google Scholar]
  11. Crotty, E. D. , Furlong L.‐A. M., and Harrison A. J.. 2024. “Ankle and Plantar Flexor Muscle‐Tendon Unit Function in Sprinters: A Narrative Review.” Sports Medicine 54, no. 3: 585–606. 10.1007/s40279-023-01967-1. [DOI] [PubMed] [Google Scholar]
  12. de Keijzer, K. L. , McErlain‐Naylor S. A., and Beato M.. 2022. “The Effect of Flywheel Inertia on Peak Power and Its Inter‐session Reliability During Two Unilateral Hamstring Exercises: Leg Curl and Hip Extension.” Frontiers in Sports and Active Living 4: 898649. 10.3389/fspor.2022.898649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. De Keijzer, K. L. , McErlain‐Naylor S. A., Iacono A. D., and Beato M.. 2020. “Effect of Volume on Eccentric Overload‐Induced Postactivation Potentiation of Jumps.” International Journal of Sports Physiology and Performance 15, no. 7: 976–981. 10.1123/ijspp.2019-0411. [DOI] [PubMed] [Google Scholar]
  14. Douglas, J. , Pearson S., Ross A., and McGuigan M.. 2018. “Kinetic Determinants of Reactive Strength in Highly Trained Sprint Athletes.” Journal of Strength & Conditioning Research 32, no. 6: 1562–1570. 10.1519/JSC.0000000000002245. [DOI] [PubMed] [Google Scholar]
  15. Ebben, W. P. , and Petushek E. J.. 2010. “Using the Reactive Strength Index Modified to Evaluate Plyometric Performance.” Journal of Strength & Conditioning Research 24, no. 8: 1983–1987. 10.1519/JSC.0b013e3181e72466. [DOI] [PubMed] [Google Scholar]
  16. Fu, K. , Chen L., Poon E. T. C., et al. 2023. “Post‐Activation Performance Enhancement of Flywheel Training on Lower Limb Explosive Power Performance.” Frontiers in Physiology 14: 1217045. 10.3389/fphys.2023.1217045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Galiano, C. , Nuñez F. J., McErlain‐Naylor S., and Floria P.. 2025. “Biomechanical Insights into Rotary Inertial Devices vs. Traditional Resistance Training in Half‐Squats: A Comparative Analysis of Force, Velocity, and Displacement.” Journal of Strength & Conditioning Research 39, no. 5: 531–539. 10.1519/JSC.0000000000005073. [DOI] [PubMed] [Google Scholar]
  18. Hall, K. B. , Immink M. A., Martin D. T., Bennett H., and Crowther R. G.. 2024. “Dose‐Response Effect of an Inertia Flywheel Postactivation Performance Enhancement Protocol on Countermovement Jump Performance.” Journal of Applied Biomechanics 40, no. 2: 147–154. 10.1123/jab.2023-0217. [DOI] [PubMed] [Google Scholar]
  19. Hobara, H. , Kanosue K., and Suzuki S.. 2007. “Changes in Muscle Activity With Increase in Leg Stiffness During Hopping.” Neuroscience Letters 418, no. 1: 55–59. 10.1016/j.neulet.2007.02.064. [DOI] [PubMed] [Google Scholar]
  20. Jarvis, P. , Turner A., Read P., and Bishop C.. 2022. “Reactive Strength Index and Its Associations With Measures of Physical and Sports Performance: A Systematic Review With Meta‐Analysis.” Sports Medicine 52, no. 2: 301–330. 10.1007/s40279-021-01566-y. [DOI] [PubMed] [Google Scholar]
  21. Kipp, K. , Kiely M., Giordanelli M., Malloy P., and Geiser C.. 2017. “The Reactive Strength Index Reflects Vertical Stiffness During Drop Jumps.” International Journal of Sports Physiology and Performance 13: 1–20. 10.1123/ijspp.2017-0021. [DOI] [PubMed] [Google Scholar]
  22. Lai, A. , Schache A. G., Brown N. A. T., and Pandy M. G.. 2016. “Human Ankle Plantar Flexor Muscle‐Tendon Mechanics and Energetics During Maximum Acceleration Sprinting.” Journal of the Royal Society Interface 13, no. 121: 20160391. 10.1098/rsif.2016.0391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Maroto‐Izquierdo, S. , Raya‐González J., Hernández‐Davó J. L., and Beato M.. 2021. “Load Quantification and Testing Using Flywheel Devices in Sports.” Frontiers in Physiology 12: 739399. 10.3389/fphys.2021.739399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Martinez‐Aranda, L. M. , and Fernandez‐Gonzalo R.. 2017. “Effects of Inertial Setting on Power, Force, Work, and Eccentric Overload During Flywheel Resistance Exercise in Women and Men.” Journal of Strength & Conditioning Research 31, no. 6: 1653–1661. 10.1519/JSC.0000000000001635. [DOI] [PubMed] [Google Scholar]
  25. McErlain‐Naylor, S. A. , and Beato M.. 2021. “Concentric and Eccentric Inertia‐Velocity and Inertia‐Power Relationships in the Flywheel Squat.” Journal of Sports Sciences 39, no. 10: 1136–1143. 10.1080/02640414.2020.1860472. [DOI] [PubMed] [Google Scholar]
  26. McKay, A. K. A. , Stellingwerff T., Smith E. S., et al. 2022. “Defining Training and Performance Caliber: A Participant Classification Framework.” International Journal of Sports Physiology and Performance 17, no. 2: 317–331. 10.1123/ijspp.2021-0451. [DOI] [PubMed] [Google Scholar]
  27. Murton, J. , Eager R., and Drury B.. 2023. “Comparison of Flywheel Versus Traditional Resistance Training in Elite Academy Male Rugby Union Players.” Research in Sports Medicine 31, no. 3: 214–227. 10.1080/15438627.2021.1954518. [DOI] [PubMed] [Google Scholar]
  28. Nagahara, R. , Mizutani M., Matsuo A., Kanehisa H., and Fukunaga T.. 2018. “Association of Sprint Performance With Ground Reaction Forces During Acceleration and Maximal Speed Phases in a Single Sprint.” Journal of Applied Biomechanics 34, no. 2: 104–110. 10.1123/jab.2016-0356. [DOI] [PubMed] [Google Scholar]
  29. Nagahara, R. , and Zushi K.. 2016. “Development of Maximal Speed Sprinting Performance With Changes in Vertical, Leg and Joint Stiffness.” Journal of Sports Medicine and Physical Fitness 57, no. 12. 10.23736/S0022-4707.16.06622-6. [DOI] [PubMed] [Google Scholar]
  30. Nishiumi, D. , Nishioka T., Saito H., Kurokawa T., and Hirose N.. 2023. “Associations of Eccentric Force Variables During Jumping and Eccentric Lower‐Limb Strength With Vertical Jump Performance: A Systematic Review.” PLoS One 18, no. 8: e0289631. 10.1371/journal.pone.0289631. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. O'Brien, J. , Browne D., Earls D., and Lodge C.. 2022. “The Efficacy of Flywheel Inertia Training to Enhance Hamstring Strength.” Journal of Functional Morphology and Kinesiology 7, no. 1: 14. 10.3390/jfmk7010014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Piqueras‐Sanchiz, F. , Sabido R., Raya‐González J., et al. 2020. “Effects of Different Inertial Load Settings on Power Output Using a Flywheel Leg Curl Exercise and Its Inter‐Session Reliability.” Journal of Human Kinetics 74, no. 1: 215–226. 10.2478/hukin-2020-0029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Puustinen, J. , Venojärvi M., Haverinen M., and Lundberg T. R.. 2023. “Effects of Flywheel vs. Traditional Resistance Training on Neuromuscular Performance of Elite Ice Hockey Players.” Journal of Strength & Conditioning Research 37, no. 1: 136–140. 10.1519/JSC.0000000000004159. [DOI] [PubMed] [Google Scholar]
  34. Raya‐González, J. , Castillo D., and Beato M.. 2021. “The Flywheel Paradigm in Team Sports: A Soccer Approach.” Strength and Conditioning Journal 43, no. 1: 12–22. 10.1519/SSC.0000000000000561. [DOI] [Google Scholar]
  35. Sabido, R. , Hernández‐Davó J. L., and Pereyra‐Gerber G. T.. 2018. “Influence of Different Inertial Loads on Basic Training Variables During the Flywheel Squat Exercise.” International Journal of Sports Physiology and Performance 13, no. 4: 482–489. 10.1123/ijspp.2017-0282. [DOI] [PubMed] [Google Scholar]
  36. Seitz, L. B. , and Haff G. G.. 2016. “Factors Modulating Post‐Activation Potentiation of Jump, Sprint, Throw, and Upper‐Body Ballistic Performances: A Systematic Review With Meta‐Analysis.” Sports Medicine 46, no. 2: 231–240. 10.1007/s40279-015-0415-7. [DOI] [PubMed] [Google Scholar]
  37. Shi, J. , Yan B., Yu M., et al. 2024. “Heavier Loads in Flywheel Exercise Induce Greater Post‐Activation Performance Enhancement in Countermovement Jumps Compared to Heavy Smith Machine Squats in Males.” Biology of Sport 41, no. 4: 231–240. 10.5114/biolsport.2024.139075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Souhaiel, M. C. , and Denis C.. 2001. “Leg Power and Hopping Stiffness: Relationship With Sprint Running Performance.” Medicine & Science in Sports & Exercise 33, no. 2: 326–333. 10.1097/00005768-200102000-00024. [DOI] [PubMed] [Google Scholar]
  39. Southey, B. , Willshire M., Connick M. J., Austin D., Spits D., and Beckman E.. 2024. “Reactive Strength Index as a Key Performance Indicator in Different Athlete Populations—A Systematic Review.” Science & Sports 39, no. 2: 129–143. 10.1016/j.scispo.2023.01.004. [DOI] [Google Scholar]
  40. Spudić, D. , Smajla D., Šarabon N., Sattler T., and Jurov I.. 2024. “Force–Velocity–Power Profiling in Flywheel Squats: Differences Between Sports and Association With Countermovement Jump and Change of Direction Performance.” Journal of Sports Sciences 42, no. 15: 1464–1476. 10.1080/02640414.2024.2394750. [DOI] [PubMed] [Google Scholar]
  41. Stratford, C. , Dos'Santos T., and McMahon J. J.. 2021. “The 10/5 Repeated Jumps Test: Are 10 Repetitions and Three Trials Necessary?” Biomechanics 1, no. 1: 1–14. 10.3390/biomechanics1010001. [DOI] [Google Scholar]
  42. Tsoukos, A. , Tsoukala M., Papadimitriou D. M., Terzis G., and Bogdanis G. C.. 2025. “Acute Effects of Low vs. High Inertia During Flywheel Deadlifts With Equal Force Impulse on Vertical Jump Performance.” Sensors 25, no. 4: 1125. 10.3390/s25041125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Weyand, P. G. , Sternlight D. B., Bellizzi M. J., and Wright S.. 2000. “Faster Top Running Speeds Are Achieved With Greater Ground Forces Not More Rapid Leg Movements.” Journal of Applied Physiology (1985) 89, no. 5: 1991–1999. 10.1152/jappl.2000.89.5.1991. [DOI] [PubMed] [Google Scholar]
  44. Wren, C. , Beato M., McErlain‐Naylor S. A., Iacono A. D., and de Keijzer K. L.. 2023. “Concentric Phase Assistance Enhances Eccentric Peak Power During Flywheel Squats: Intersession Reliability and the Linear Relationship Between Concentric and Eccentric Phases.” International Journal of Sports Physiology and Performance 18, no. 4: 428–434. 10.1123/ijspp.2022-0349. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S1: Sphericity assessment and repeated‐measures ANOVA results.

EJSC-26-e70264-s002.docx (28.6KB, docx)

Table S2: Linear mixed‐effects model estimates and diagnostics.

EJSC-26-e70264-s001.docx (36.5KB, docx)

Data Availability Statement

The data that support the findings of this study are not publicly available owing to privacy and ethical restrictions concerning the athlete participants; however, they can be requested from the corresponding author upon reasonable request.

Articles from European Journal of Sport Science are provided here courtesy of Wiley

RESOURCES