Multi-joint isokinetic strength profiling as a predictor of vertical jump performance in elite freestyle wrestlers: A cross-sectional principal component analysis

Abstract

Wrestling requires coordinated lower-limb strength for explosive techniques. While isokinetic testing isolates joint-specific strength and countermovement jumps assess functional power, comprehensive multi-joint assessments with advanced factor analysis remain unexplored in elite wrestlers. Twenty-one national-level male freestyle wrestlers (22.3 ± 2.8 years; 76.2 ± 14.7 kg) completed cross-sectional neuromuscular assessments. Isokinetic concentric strength testing at 60°/s evaluated hip, knee, and ankle joint functions bilaterally using an Isomed 2000 dynamometer. Vertical jump power was assessed through countermovement jumps under bilateral and unilateral conditions using the Optojump system. All strength and power variables were normalized to fat-free mass determined via air displacement plethysmography. Statistical analyses included Pearson correlations with Benjamini–Hochberg correction, stepwise multiple regression, and principal component analysis. Bilateral jump performance correlated strongly with knee extension (R = 0.798, P < .001) and hip extension strength (R = 0.664, P = .001), with combined predictors explaining 68.8% of variance (F₂,₁₈ = 19.8, P < .001). Principal component analysis identified a primary extensor factor (hip and knee extension, 52.6% variance) and an auxiliary factor (ankle plantar flexion and hip adduction, 19.0% variance). Unilateral jump models demonstrated lower explanatory power (dominant: 39.8%, non-dominant: 25.7%). Multi-joint isokinetic strength assessment, particularly hip and knee extension capacity, provides a robust prediction of vertical jump performance in elite wrestlers. The identified primary extensor strength factor represents a fundamental neuromuscular capacity underlying explosive lower-limb function, supporting targeted strength development strategies for wrestling-specific performance enhancement.

1. Introduction

Wrestling is a high-intensity intermittent sport that requires a complex interaction of strength and power during repeated explosive actions, including throwing techniques, forceful opponent manipulation, and rapid postural recovery from defensive situations. Body control and the ability to rapidly generate force during unilateral or bilateral lower-limb actions are critical for successful performance.^[1] To perform these actions, lower-limb strength and power are considered critical determinants during both offensive and defensive phases, which require high neuromuscular output.^[2,3] Therefore, lower-limb assessments are widely used in wrestling to evaluate physical performance. The biomechanical foundation of vertical jumping involves sequential activation of hip, knee, and ankle extensors in a coordinated kinetic chain, with hip extension contributing 23% to 28% of total jump power, knee extension providing 49% to 56%, and ankle plantarflexion accounting for 15% to 18%.^[4,5] Specifically, monoarticular muscles are primarily responsible for producing force at individual joints, and bi-articular muscles, such as the hamstrings, gastrocnemius, and rectus femoris, play a critical role in the coordination and energy transfer between adjacent joints.^[6] This multi-joint coordination pattern underlies the rationale for comprehensive strength assessment across all major lower-limb joints. In a meta-analysis of studies conducted between 2010 and 2021, the countermovement jump (CMJ) was reported as the most frequently used method for assessing the lower extremity explosive strength in wrestlers, followed by the long jump and squat jump.^[7] Considering the high cost and limited accessibility of isokinetic testing compared to field-based assessments such as the CMJ, examining the relationship between these 2 methods is of particular importance for the development of practical and cost-effective performance monitoring strategies in wrestling.^[7–9]

Isokinetic strength tests and CMJ assessments differ fundamentally in their biomechanical characteristics, which may influence their interrelationship. Isokinetic testing is performed in an open kinetic chain, isolating single-joint actions at constant angular velocities, thereby limiting intersegmental coordination and energy transfer.^[10,11] In contrast, CMJ is a closed kinetic chain movement that involves the coordinated action of multiple joints and muscle groups, allowing for stretch-shortening cycle dynamics, elastic energy reutilization, and intersegmental force transmission.^[4,12] These biomechanical distinctions are reflected in empirical findings indicating that isokinetic and CMJ tests assess different neuromuscular capacities. For example, factor analysis by Menzel et al demonstrated that isokinetic torque values and CMJ performance load on separate components, suggesting that these assessments are not functionally interchangeable.^[11] The discrepancies are likely due to differences in angular velocity profiles (constant vs variable), muscle involvement (isolated vs multi-joint), and movement specificity (open vs closed kinetic chain).^[13,14]

Despite these methodological distinctions, there is a general consensus in the literature that the quadriceps femoris as a knee extensor muscle group plays a crucial role in jumping ability.^[9,15–18] Previous research examining single-joint relationships has reported moderate associations (R = 0.40–0.65), while comprehensive multi-joint assessments remain largely unexplored in wrestling populations, representing a significant gap in understanding coordinated strength–power relationships.^[9,19,20] Particularly when torque values are normalized to body mass or fat-free mass (FFM) and measured at moderate to high angular velocities.^[9,20] However, other studies have failed to observe such relationships, particularly when using absolute torque values without normalization for FFM or when focusing on single-joint assessments rather than comprehensive multi-joint evaluations.^[5,21,22] These discrepancies highlight a lack of consensus in the literature and underscore the need for further investigation into the contextual and methodological factors that may influence the strength of this association. Principal component analysis (PCA) has emerged as a valuable tool for^[23 identifying underlying strength factor structures yet its application to multi-joint isokinetic profiling in wrestlers remains unexplored. PCA provides a data reduction technique to identify underlying strength factor structures that may better explain functional movement patterns than individual joint assessments.^[24]

Therefore, this study aims to address the conflicting results in the current literature by comprehensively examining the relationship between isokinetic strength and CMJ performance in elite wrestlers, a population in which lower-limb strength plays a pivotal role in athletic success. Unlike previous studies that have often focused solely on the knee joint or bilateral jumps, the present research employs comprehensive multi-joint isokinetic assessments (hip, knee, and ankle), evaluates both bilateral and unilateral jump performance, and applies FFM normalization with advanced statistical modeling including PCA, as measured by the gold-standard method of air displacement plethysmography, to account for individual differences in body composition. By integrating these methodological strengths, the study aims to provide a more comprehensive understanding of how joint-specific strength capacities relate to functional power output in sport-specific contexts. Air displacement plethysmography was selected for FFM determination based on its validated accuracy for body composition assessment in athletic populations.^[25,26] We hypothesized that: isokinetic hip and knee extension strength would demonstrate strong positive correlations with vertical jump power output; bilateral jump performance would show stronger associations with isokinetic strength measures than unilateral conditions due to enhanced mechanical advantage and reduced balance constraints; and combined multi-joint strength variables would explain a substantial proportion of jump performance variance through identifiable strength factor structures. As a secondary objective, we examined inter-limb strength asymmetries and their relationship with unilateral jump performance, given their potential relevance for injury risk assessment and performance optimization in wrestling.

2. Materials and methods

This article focuses on understanding the relationship between lower-limb anaerobic power and joint strength. Specifically, the relationship between relative vertical jump power, calculated as power multiplied by FFM, and relative isokinetic strength of the hip, knee, and ankle joints, expressed as a ratio of torque (in Newton-meters) to FFM, is examined. The study examines this relationship using data collected from national-level wrestlers.

2.1. Participants

Twenty-one male wrestlers, all competing at the national level in the freestyle discipline, volunteered to participate in this study. An a priori power analysis using G*Power 3.1.9.7 was conducted to determine an adequate sample size for detecting medium-to-large effect sizes (R = 0.50) in correlation analyses with 80% power and α = 0.05, indicating a minimum requirement of 20 participants.^[27] Our sample of 21 elite wrestlers exceeds this threshold and is consistent with similar isokinetic-jump relationship studies in elite combat sport populations (n = 15–25).^[28–30] Their mean (±standard deviation) characteristics were as follows: age 22.3 (±2.7) years, height 173.8 (±5.1) cm, weight 76.2 (±14.6) kg, FFM 66.7 (± 8.9) kg, and 10.1 (± 3.39) years of wrestling experience (Table 1). Before participation, all subjects read and signed an informed consent form.

Table 1.

Comprehensive participant characteristics, anthropometric measures, and performance variables (n = 21).

	Variable	Mean ± SD	95% CI	Range	CV (%)
Demographic characteristics	Age (yr)	22.3 ± 2.8	[21.0–23.6]	18.0–28.0	12.6
	Height (cm)	173.8 ± 5.2	[171.4–176.2]	165.0–185.0	3
	Body mass (kg)	76.2 ± 14.7	[69.5–82.9]	58.0–110.0	19.3
	Training experience (yr)	10.1 ± 3.2	[8.6–11.6]	5.0–17.0	31.7
	Weekly training volume (h)	18.4 ± 3.1	[17.0–19.8]	12.0–24.0	16.8
Body composition	Fat percentage (%)	11.5 ± 5.1	[9.2–13.8]	4.2–22.1	44.3
	Fat-free mass (kg)	66.7 ± 8.9	[62.6–70.8]	52.1–81.3	13.3
	Body mass index (kg/m²)	25.2 ± 4.1	[23.3–27.1]	19.8–32.1	16.3
Jump performance (watts/FFM)	Double-leg CMJ	1145.8 ± 161.4	[1072.3–1219.2]	847.2–1416.8	14.1
	Dominant single-leg CMJ	678.4 ± 102.5	[631.7–725.0]	440.6–848.5	15.1
	Non-dominant single-leg CMJ	682.7 ± 115.2	[630.2–735.1]	440.6–853.9	16.9
Bilateral isokinetic strength (N/FFM)	Hip extension	9.84 ± 1.65	[9.09–10.59]	6.38–12.93	16.8
	Hip flexion	3.77 ± 0.49	[3.55–4.00]	2.68–4.53	13
	Hip abduction	2.76 ± 0.45	[2.56–2.97]	1.86–3.51	16.3
	Hip adduction	4.66 ± 0.84	[4.28–5.04]	3.09–6.15	18
	Knee extension	6.16 ± 0.86	[5.77–6.55]	4.77–7.92	14
	Knee flexion	4.00 ± 0.54	[3.75–4.24]	2.73–4.88	13.5
	Ankle plantar flexion	3.07 ± 0.48	[2.85–3.29]	2.00–3.73	15.6
	Ankle dorsiflexion	1.71 ± 0.22	[1.61–1.81]	1.22–2.05	12.9
Medical history	Competition level	National (n = 21)	–	–	–
	Lower-limb injury (past 12 mo)	None	–	–	–
	Major joint surgery (lifetime)	None	–	–	–
	Current rehabilitation status	None required	–	–	–

CI = confidence interval, CMJ = countermovement jump, CV = coefficient of variation, FFM = fat-free mass, N = Newton, SD = standard deviation.

2.1.1. Inclusion criteria

Participants were required to: be actively involved in wrestling training for at least the past 12 months; have no musculoskeletal disorders affecting lower extremity function; and be able to complete both bilateral and unilateral counter movement jump assessments, as well as the isokinetic strength testing protocol.

2.1.2. Exclusion criteria

Participants were excluded if they: had sustained a recent injury to the hip, knee, or ankle joints; presented with any pathology preventing safe participation in testing procedures; had undergone knee surgery or received intra-articular cortico-steroid injections within the last 6 months; or were unable to perform double- or single-leg vertical jump assessments.

2.2. Procedures

Participants completed 2 separate laboratory sessions. During the initial visit, body fat percentage was assessed, followed by the administration of vertical jump performance tests. In the 2nd session, which was scheduled 24 hours later, isokinetic strength assessments were conducted. To minimize the potential confounding effects of fatigue, all participants were instructed to abstain from any form of training or strenuous physical activity for at least 48 hours before each testing session. All assessments were conducted between 10:00 and 12:00 am under controlled environmental conditions, with the ambient temperature maintained between 22.5 and 24.0°C. Adequate hydration status was ensured for all participants throughout the testing procedures.

2.2.1. Vertical countermovement jump power

The repeated CMJ (CMJ5) test, consisting of 5 jumps, was used to measure the explosive strength of the lower extremities. The CMJ test is the most reliable and valid field test for assessing lower-limb explosive strength.^[8,31] This test is suitable for evaluating lower extremity power in wrestlers.^[7]

Before the performance tests, all subjects completed a 10-minute warm-up on a stationary bicycle ergometer, followed by activation and dynamic stretching of the hip, knee, and ankle muscles. Before the jump tests, the athletes performed single-leg and double-leg jumps to become accustomed to the movement.

The CMJ5 tests were conducted using the Optojump system (Microgate, Bolzano, Italy), which comprises 2 parallel bars (a transmitting unit and a receiving unit) aligned 92 cm apart from each other. The Optojump bars were connected to a personal computer, and the Microgate software (Optojump software, version 3.01.0001) was used to quantify the bounce force. Five consecutive jumps were performed on each condition (double-leg countermovement jump, dominant-leg countermovement jump, non-dominant-leg countermovement jump) with a 60-second recovery between trials. The 5CMJ protocol demonstrates high intra-session reliability (intraclass correlation coefficient [ICC] = 0.94–0.97, CV = 1.57–1.88%) for flight time and relative mean force in similar populations.^[8,30,31] The jumps were performed from an upright position with hands on hips to prevent an increase in eccentric load and elastic force due to arm movement.^[31] Participants were instructed to jump as fast and as high as possible, which was encouraged by teaching them to achieve the highest power output. A self-selected countermovement depth was chosen.^[32] Participants were required to land at the same point as take-off and to bounce with their legs extended on landing to avoid knee flexion and altering measurements.^[31] For this purpose, the jumps were recorded and the landing phases checked. Series with incorrect landings were repeated.

The performance of the last 4 jumps was recorded, as the 1st jump did not include a countermovement. The jump that produced the highest power of the 4 jumps (in each series) was recorded for subsequent analysis. The OptoJump system, combined with the Microgate interface, calculates the maximum power output during jumps using contact time and flight time parameters, which exhibit near-perfect reliability and correlate well with force platforms.^[33] The OptoJump system has demonstrated an excellent ICC in predicting vertical jump height compared to a force platform (ICC = 0.99, 95% confidence interval: 0.97–0.99; P < .001).^[33,34] In addition to its high reliability, the OptoJump power calculation formula (P = g².T_f.(T_f + T_c)/4.T_c) has been validated against force platform measurements with near-perfect correlation (R = 0.99, 95% confidence interval: 0.98–0.99) and provides reliable power estimates through flight time-contact time relationships.^[33,35]

The formula used by the Microgate software to estimate the maximum power is:

$${\displaystyle \begin{array}{l}{\displaystyle {\displaystyle \overline{P=g^2\cdot T_{\mathrm{f}}\cdot \frac{T_{\mathrm{f}}+T_{\mathrm{c}}}{4\cdot T_{\mathrm{c}}}}}}\end{array}}$$

g = gravity acceleration (9.81 m/s²), T_f = flight time, T_c = contact time.

A short contact time in combination with a longer flight time is an indication of an above-average neuromuscular power output. The measured power was normalized to the FFM and expressed in watts. Power output was calculated using the OptoJump formula (P = g²·T_f·(T_f + T_c)/(4·T_c)), where g represents gravitational acceleration (9.81 m/s²), T_f is flight time, and T_c is contact time. All power values were normalized to FFM rather than body mass to account for the specific contribution of lean tissue to force production.^[36]

2.2.2. Isokinetic strength of lower body joints

An Isomed 2000 isokinetic dynamometer (Lumex Corporation, Ronkonkoma) was used to measure isokinetic concentric–concentric torque developed during hip flexion–extension, hip abduction–adduction, knee extension–knee flexion, and ankle plantar flexion–ankle dorsiflexion. The validity and reliability of the device, specifically for lower-limb kinematics, have been demonstrated in previous studies.^[37,38] Two physiotherapists with 5 years of experience performed all the isokinetic tests. All isokinetic tests were performed on both legs separately to compare the results with unilateral jumping variables. The order of the tests was the same for all subjects.

For the isokinetic tests, participants were positioned on the seat of the isokinetic dynamometer. They performed 10 repetitions of concentric contractions at a velocity of 90°/s as a warm-up, followed by a 120-second rest period. The warm-up on the isokinetic machine was chosen to improve specificity and familiarity with the following test. The participant then performed 5 concentric repetitions (velocity: 60°/s and range of motion: test-specific) at maximal effort while constantly receiving standardized verbal encouragement: “faster.”^[30] A standardized rest period of 120 seconds was provided between the assessments of each limb.^[30] The positioning of the participants on the isokinetic device and the alignment of the dynamometer with the joint axis were carried out as specified in the manufacturer’s operating instructions and are detailed below.

2.2.3. Hip extension and hip flexion

This measurement was performed in the supine position, ensuring that both hip and knee joints were in a neutral position of 0°. The starting point was set at the neutral position of the leg. The axis of rotation of the hip joint in the sagittal plane (greater trochanter) was aligned with the axis of rotation of the dynamometer. The lever arm length was adjusted to the length of the subject’s thigh and attached to the distal femur. A belt was fastened around the pelvis to prevent it from lifting off the bench during the hip extension movements. Additionally, shoulder pads were used to immobilize the participant’s shoulders, ensuring they could not slide on the bench and had a stable point for generating torque against the lever arm. The non-tested leg rested passively on the bench and was not allowed to support the movement of the tested leg. The range of motion assessed for hip flexion ranged from 10° to 100°.^[39]

2.2.4. Hip abduction and hip adduction

The assessment of hip abduction (moving the leg away from the midline) and hip adduction (moving the leg towards the midline) strength was performed in a supine position, ensuring that both hip and knee joints were in a neutral 0°-position. The starting point was set at the neutral position of the leg. Participants lie on their right side to test the left hip muscles, and vice versa on the left side to test the right hip muscles. The axis of rotation of the hip joint in the frontal plane (femoral head) was aligned with the axis of rotation of the device and adjusted to the individual pelvic height. The lever arm length of the test device was adjusted to the length of the femur and attached to the lower (distal) part of the femur. The lower hip and leg (the non-tested side) were slightly bent, and the leg was attached to the bench at the femur. This grounding of the leg and body helped prevent the body from lifting, even in strong participants. Additionally, a belt was strapped around the pelvis to restrict any compensatory movements further. The range of motion tested for hip abduction ranged from 0° to 60°.^[39]

2.2.5. Knee extension and knee flexion

This test was performed while the subject was in a sitting position. Initially, the seat length and adapter length were adjusted to fit the thigh and lower leg, respectively. To effectively transfer the force generated by the thigh muscles to the resistance pad, the adapter was attached to the distal part of the lower leg, slightly above the ankle. The axis of rotation of the knee joint was then aligned with the axis of rotation of the dynamometer. The backrest and seat were tilted slightly to achieve a hip angle of approximately 90°.^[39] The starting point of the knee was set to full extension. To minimize any compensatory movements, the test subject’s thigh bone was secured with straps, the pelvis was stabilized with a belt, and the shoulders were fixed with shoulder pads. The tested range of motion for knee flexion ranged from 5° to 90°.^[39]

2.2.6. Ankle plantarflexion and dorsiflexion

This strength measurement was performed in the supine position. First, the subject’s foot was placed on the platform of an adapter specifically designed to assess the strength of plantar flexion and dorsiflexion of the ankle. The length of the adapter was adjusted to the length of the subject’s lower leg. The pivot point of the ankle joint (malleolus lateralis) was then aligned with the axis of rotation of the dynamometer. The knee joint was flexed to approximately 90° (measured with a goniometer) while the foot remained in position on the adapter.^[39] In this arrangement, the hip on the tested side was flexed by approximately 30°, while the opposite hip was in a neutral position. To prevent the subject from exerting force on the foot platform through leg extension, which could influence the results of the plantar flexion force, an additional support bar was placed behind the tested leg. To minimize compensatory movements, the subject’s foot, ankle, and thigh were secured with straps. The pelvis was stabilized with a belt, and the shoulders were fixed with shoulder pads. The range of motion tested ranged from −20° of dorsiflexion to 35° of plantarflexion. The 0°-position for the ankle joint was defined as a neutral position in which the joint was neither plantarflexed nor dorsiflexed.^[39]

2.2.7. FFM measurement

The participants’ body fat percentage was estimated using air displacement plethysmography. The BOD POD device was used to measure body composition and calculate body fat percentage, which in turn was used to determine FFM.^[40] The BOD POD device measures the air volume of the chamber with and without the subject seated in the test chamber. The difference between these measurements gives the subject’s body volume. The ICC of the BOD POD was ICC = 0.991.^[41] In a study involving a total of 66 young male national wrestlers, no significant differences were found between the BOD POD and hydrostatic weighing (the reference method) for density, body fat, and FFM. The standard errors of the estimates for body fat, as determined by the BOD POD with hydrostatic weighing, were 2.12% (hydrated) and 2.16% (dehydrated); the prediction errors were 2.35% (hydrated) and 2.49% (dehydrated) in both hydrated and dehydrated conditions.^[42]

All participants wore minimal, tight-fitting clothing, including Lycra swimsuits and caps, to minimize hair interference. The average air volume in the lungs and thorax during normal breathing, as measured using the standard BOD POD plethysmograph technique,^[40] was adjusted. Body density was calculated by dividing body mass by body volume, which was used to determine body fat percentage using Siri (1961) equation. The FFM values obtained were used to calculate “watts” for jumping and “relative peak force” for isokinetic measurements.

2.3. Statistical analysis

Statistical analyses were performed using Statistical Package for the Social Sciences version 29.0 (IBM Corporation, Armonk) and R version 4.3.0 (R Foundation for Statistical Computing, Vienna, Austria) 4.3.0 for advanced modeling. Given the specialized elite athlete population and comprehensive multi-joint assessment, our sample of 21 wrestlers was deemed sufficient for the primary research objectives. Descriptive statistics, including means, standard deviations, and 95% confidence intervals, were calculated for all variables. Data normality was assessed via Shapiro–Wilk tests, Q–Q plots, histograms, and skewness/kurtosis values (acceptable range: ±2.0). Homoscedasticity was evaluated using Levene test, and outliers were identified using the interquartile range method (IQR × 1.5) and standardized z-scores (∣| z∣|>3.29). Sample size adequacy was determined a priori through power analysis rather than post hoc evaluation, ensuring sufficient statistical power for detecting theoretically meaningful effect sizes independent of observed results.

Pearson product–moment correlations were used to examine bivariate relationships between normalized isokinetic strength and jump performance measures. The Benjamini–Hochberg false discovery rate correction was applied (controlled rate of 0.05) to address concerns about multiple comparisons. Effect sizes were interpreted using Cohen conventions: small (R = 0.10), medium (R = 0.30), and large (R = 0.50).^[27] Precision estimates for correlation coefficients were provided via 95% confidence intervals calculated using Fisher z-transformation.

Hierarchical stepwise multiple regression analyses were conducted to identify and quantify predictors of bilateral, dominant-leg, and non-dominant-leg CMJ power outputs. Independent variables included normalized isokinetic strength measures, which exhibited significant bivariate correlations (P < .05). The forward selection method was employed, using entry and removal criteria of P ≤ .05 and P ≥ .10, respectively. Model assumptions, including linearity, residual independence (Durbin–Watson test), homoscedasticity (Breusch–Pagan test), and multicollinearity (variance inflation factor < 5.0, tolerance > 0.20), were rigorously evaluated.^[24]

Comparative analyses explored strength differences and their contribution to jumping performance. Normalized isokinetic strength values were categorized into tertiles (low, moderate, high) for each muscle group. One-way analyses of variance with post hoc Tukey HSD tests were used to compare jump performance across strength tertiles. Effect sizes were calculated using partial eta-squared (ηp²), interpreted as small (0.01), medium (0.06), and large (0.14) effects.

PCA was conducted on normalized isokinetic strength variables to identify underlying strength factors. Sampling adequacy was verified using the Kaiser–Meyer–Olkin measure (target ≥ 0.60), and Bartlett test of sphericity confirmed the suitability of factor analysis. Components with eigenvalues >1.0 were retained, and a varimax rotation was applied to enhance interpretability.

Limb asymmetry indices were calculated as: [(dominant − non-dominant)/dominant] × 100, with asymmetries > 10% considered potentially performance-limiting. Paired-samples t tests examined between-limb differences in isokinetic strength, and correlation analysis assessed their relationships with unilateral jump performance asymmetries.

Model validation employed leave-one-out cross-validation to assess the stability and generalizability of the regression model. Predictive accuracy was evaluated using the root mean square error of prediction and coefficient of determination for prediction (Q2). All statistical assumptions were verified, and appropriate non-parametric alternatives were utilized when assumptions were violated. Statistical significance was set at α = 0.05, with effect sizes reported alongside P-values to provide practical relevance. Missing data (<2%) were handled via listwise deletion after confirming data were missing completely at random (Little Missing Completely at Random test).

3. Results

3.1. Participant characteristics and descriptive statistics

Twenty-one elite male freestyle wrestlers completed all testing procedures. All measured variables demonstrated normal distribution characteristics (skewness and kurtosis values within ± 2.0), satisfying the assumptions for parametric statistical analyses. The Shapiro–Wilk test confirmed normality for all variables except non-dominant knee flexion strength (W = 0.892, P = .031), which was subsequently analyzed using appropriate robust statistical methods (Table 2).

Table 2.

Descriptive statistics for countermovement jump performance and isokinetic strength variables normalized to fat-free mass.

	Variable	Mean ± SD	95% CI	Range	CV (%)
Jump performance (watts)	Double-leg CMJ	1145.78 ± 161.37	[1072.33–1219.23]	847.20–1416.80	14.1
	Dominant single-leg CMJ	678.37 ± 102.53	[631.70–725.04]	440.58–848.50	15.1
	Non-dominant single-leg CMJ	682.68 ± 115.20	[630.24–735.12]	440.58–853.93	16.9
Bilateral isokinetic strength (N/FFM)	Hip extension	9.84 ± 1.65	[9.09–10.59]	6.38–12.93	16.8
	Hip flexion	3.77 ± 0.49	[3.55–4.00]	2.68–4.53	13
	Hip abduction	2.76 ± 0.45	[2.56–2.97]	1.86–3.51	16.3
	Hip adduction	4.66 ± 0.84	[4.28–5.04]	3.09–6.15	18
	Knee extension	6.16 ± 0.86	[5.77–6.55]	4.77–7.92	14
	Knee flexion	4.00 ± 0.54	[3.75–4.24]	2.73–4.88	13.5
	Ankle plantar flexion	3.07 ± 0.48	[2.85–3.29]	2.00–3.73	15.6
	Ankle dorsiflexion	1.71 ± 0.22	[1.61–1.81]	1.22–2.05	12.9
Unilateral isokinetic strength-dominant limb (N/FFM)	Hip extension	4.94 ± 0.85	[4.55–5.32]	3.14–6.60	17.2
	Hip flexion	1.91 ± 0.25	[1.79–2.02]	1.31–2.27	13.1
	Hip abduction	2.26 ± 0.46	[2.06–2.47]	1.44–3.08	20.4
	Hip adduction	1.35 ± 0.27	[1.22–1.47]	0.92–1.81	20
	Knee extension	3.10 ± 0.49	[2.88–3.33]	2.04–4.16	15.8
	Knee flexion	2.06 ± 0.25	[1.95–2.17]	1.61–2.41	12.1
Unilateral isokinetic strength-non-dominant limb (N/FFM)	Hip extension	4.90 ± 0.91	[4.49–5.31]	3.24–7.21	18.6
	Hip flexion	1.87 ± 0.30	[1.73–2.00]	1.38–2.51	16
	Hip abduction	2.40 ± 0.48	[2.18–2.62]	1.64–3.50	20
	Hip adduction	1.42 ± 0.21	[1.32–1.51]	0.94–1.72	14.8
	Knee extension	3.06 ± 0.53	[2.82–3.30]	2.00–3.93	17.3
	Knee flexion	1.94 ± 0.31	[1.79–2.08]	1.04–2.47	16
Jump temporal characteristics	Bilateral contact time (s)	0.35 ± 0.08	[0.31–0.39]	0.22–0.51	22.9
	Bilateral flight time (s)	0.52 ± 0.06	[0.49–0.55]	0.41–0.63	11.5
	Bilateral RSI	1.64 ± 0.32	[1.49–1.79]	1.12–2.27	19.5
	Dominant contact time (s)	0.41 ± 0.09	[0.37–0.45]	0.28–0.58	22
	Dominant flight time (s)	0.46 ± 0.05	[0.44–0.48]	0.37–0.55	10.9
	Dominant RSI	1.18 ± 0.28	[1.05–1.31]	0.71–1.72	23.7
	Non-dominant contact time (s)	0.43 ± 0.10	[0.38–0.48]	0.29–0.62	23.3
	Non-dominant flight time (s)	0.45 ± 0.05	[0.43–0.47]	0.36–0.54	11.1
	Non-dominant RSI	1.12 ± 0.26	[1.00–1.24]	0.68–1.61	23.2

CI = confidence interval, CMJ = countermovement jump, CV = coefficient of variation, FFM = fat-free mass, N = Newton, RSI = Reactive Strength Index, SD = standard deviation.

Jump performance metrics demonstrated the following temporal characteristics: contact times of 0.35 ± 0.08 seconds (bilateral), 0.41 ± 0.09 seconds (dominant limb), and 0.43 ± 0.10 seconds (non-dominant limb); flight times of 0.52 ± 0.06 seconds (bilateral), 0.46 ± 0.05 seconds (dominant limb), and 0.45 ± 0.05 seconds (non-dominant limb). Reactive strength indices, calculated as the ratio of flight time to contact time, were 1.64 ± 0.32 (bilateral), 1.18 ± 0.28 (dominant limb), and 1.12 ± 0.26 (non-dominant limb), indicating superior neuromuscular efficiency in bilateral compared to unilateral conditions.

3.2. Correlation analysis between isokinetic strength and jump performance

Pearson product–moment correlations with Benjamini–Hochberg false discovery rate correction revealed distinct patterns of association between joint-specific strength measures and vertical jump performance across different movement conditions (Table 3).

Table 3.

Correlation coefficients between isokinetic strength variables and countermovement jump performance with FDR correction.

Strength variable	Double-leg CMJ r [95% CI]	Dominant single-leg CMJ r [95% CI]	Non-dominant single-leg CMJ r [95% CI]
Hip extension	0.664*** [0.326 to 0.852]	0.569** [0.182 to 0.803]	0.507* [0.146 to 0.713]
Knee extension	0.798*** [0.559 to 0.915]	0.578** [0.195 to 0.808]	0.432 [-0.142 to 0.822]
Hip flexion	0.341 [-0.106 to 0.674]	0.491* [0.152 to 0.757]	0.315 [-0.132 to 0.652]
Knee flexion	0.527* [0.123 to 0.781]	0.491* [0.150 to 0.760]	0.291 [-0.158 to 0.634]
Hip abduction	0.312 [-0.135 to 0.651]	0.427 [0.021 to 0.705]	0.358 [-0.090 to 0.678]
Hip adduction	0.401 [-0.035 to 0.702]	0.438* [0.034 to 0.670]	0.298 [-0.151 to 0.640]
Ankle plantar flexion	0.224 [-0.214 to 0.587]	0.168 [-0.275 to 0.557]	0.287 [-0.162 to 0.631]
Ankle dorsiflexion	0.289 [-0.160 to 0.632]	0.201 [-0.243 to 0.574]	0.095 [-0.338 to 0.491]

CI = confidence interval, CMJ = countermovement jump. Bold values indicate statistically significant correlations after FDR correction.

^*FDR-corrected significance level: P < .05.

^**FDR-corrected significance level: P < .01.

^***FDR-corrected significance level: P < .001.

For bilateral jumping, knee extension strength demonstrated the strongest association (R = 0.798, P < .001), followed by hip extension strength (R = 0.664, P = .001). These 2 variables survived FDR correction and exhibited large effect sizes according to Cohen criteria. Moderate correlations were observed for knee flexion strength (R = 0.527, P = .014), though this association did not maintain significance following multiple comparison correction.

In unilateral jumping conditions, both dominant and non-dominant limbs showed significant correlations with hip and knee extension strength, albeit with reduced magnitude compared to bilateral performance. The dominant limb demonstrated stronger associations across multiple muscle groups, including hip adduction (R = 0.438, P = .045) and hip flexion strength (R = 0.491, P = .023). Post hoc power analysis confirmed adequate statistical power (1 - β > 0.95) for detecting the observed large effect sizes between knee extension strength (R = 0.798) and hip extension strength (R = 0.664) with bilateral jump performance, validating the a priori sample size determination.

3.3. Stepwise multiple regression analysis

Hierarchical stepwise regression analysis identified the most significant predictors of jump performance across different movement conditions, with models constructed using only variables that demonstrated significant bivariate correlations (Table 4).

Table 4.

Stepwise multiple regression models predicting countermovement jump performance.

Condition	Model statistics	Predictors	β	SE	t	P
Double-leg CMJ	R² = 0.688, Adj R² = 0.653, F(2,18) = 19.8, P < .001	Knee extension	0.512	0.128	4.01	.001
Double-leg CMJ	R² = 0.688, Adj R² = 0.653, F(2,18) = 19.8, P < .001	Hip extension	0.394	0.128	3.08	.007
Dominant single-leg CMJ	R² = 0.398, Adj R² = 0.331, F(2,18) = 5.94, P = .010	Hip extension	0.402	0.156	2.58	.019
Dominant single-leg CMJ	R² = 0.398, Adj R² = 0.331, F(2,18) = 5.94, P = .010	Knee extension	0.318	0.156	2.04	.057
Non-dominant single-leg CMJ	R² = 0.257, Adj R² = 0.218, F(1,19) = 6.57, P = .019	Hip extension	0.507	0.184	2.76	.013

Adj = adjusted, SE = standard error, β = standardized regression coefficient.

The bilateral jumping model explained 68.8% of the variance in performance, with knee extension strength emerging as the primary predictor (β = 0.512, P = .001), followed by hip extension strength (β = 0.394, P = .007). Cross-validation using leave-one-out methodology yielded a prediction coefficient of Q² = 0.634, indicating robust model stability.

For unilateral jumping, the dominant limb model accounted for 39.8% of variance, with hip extension strength as the primary predictor (β = 0.402, P = .019). The non-dominant limb showed reduced explanatory power (25.7% variance), with hip extension strength as the sole significant predictor (β = 0.507, P = .013).

3.4. Inter-limb asymmetry analysis

Paired-samples t tests revealed no significant between-limb differences in isokinetic strength measures (all P > .05). However, individual asymmetry indices exceeded the 10% threshold considered potentially performance-limiting in several participants (Table 5).

Table 5.

Inter-limb asymmetry indices for strength and jump performance variables.

Variable	Mean asymmetry (%)	95% CI	Athletes > 10% asymmetry (n)
Hip extension strength	2.8 ± 12.4	[-3.0 to 8.6]	7 (33.3%)
Knee extension strength	1.2 ± 15.8	[-6.1 to 8.5]	6 (28.6%)
Jump performance	-0.6 ± 9.7	[-5.0 to 3.8]	5 (23.8%)

Hip extension strength asymmetry correlated moderately with jump performance asymmetry (R = 0.441, P = .042), suggesting that strength imbalances may contribute to functional performance differences between limbs.

3.5. Strength tertile analysis

Strength tertile analysis revealed significant performance gradations across hip extension strength levels. Wrestlers in the highest tertile (>10.8 N/FFM) achieved 1285.6 ± 118.9 watts in bilateral jumps, significantly exceeding both moderate tertile (1126.3 ± 142.1 watts) and low tertile performance (1025.4 ± 153.2 watts; F(2,18) = 6.24, P = .009, ηp² = 0.41, post hoc P = .007 between high-low groups). Similarly, knee extension strength tertiles demonstrated significant performance differences (F(2,18) = 11.82, P < .001, ηp² = 0.57), with large effect sizes indicating substantial practical significance for both muscle groups.

Similarly, knee extension strength tertiles showed significant performance gradations (F(2,18) = 11.82, P < .001, ηp² = 0.57), with effect sizes indicating large practical significance. These findings support the threshold-dependent relationship between joint-specific strength and functional power output in elite wrestlers.

3.6. PCA of strength variables

PCA identified 2 principal components explaining 71.3% of the total variance in isokinetic strength measures. Component 1 (eigenvalue = 4.2, 52.5% variance) loaded heavily on hip and knee extension variables (loadings > 0.70), representing a “primary extensor strength” factor (Fig. 1A). The standardized loading pattern demonstrated that hip extension, knee extension, and hip flexion contributed most strongly to this component. In contrast, ankle plantar flexion and hip adduction defined component 2 (eigenvalue = 1.5, 18.8% variance), representing “auxiliary strength” capacity (Fig. 1B). Eigenvalue decomposition confirmed retention of 2 components based on the elbow criterion, with subsequent components contributing minimal additional variance (Fig. 1C). Variable correlation analysis revealed that primary extensors clustered near the unit circle perimeter, indicating strong component contributions, while auxiliary variables showed moderate loadings with greater independence (Fig. 1D). The primary extensor strength component correlated strongly with bilateral jump performance (R = 0.763, P < .001) and moderately with unilateral performance (R = 0.524, P = .014), reinforcing the importance of coordinated hip–knee extensor function in vertical jumping tasks.

Figure 1.

Principal component analysis of isokinetic strength variables in elite wrestlers. (A) Biplot showing individual wrestler positioning (points) according to strength profiles on PC1–PC2 axes, with color gradient representing bilateral countermovement jump power (watts). Variable vectors indicate strength measure contributions to the component structure. (B) Standardized component loadings for 8 strength variables across PC1 (primary extensor strength, 52.6% variance) and PC2 (auxiliary strength capacity, 19.0% variance). Dashed horizontal lines demarcate ± 0.70 loading threshold for statistical interpretation. (C) Eigenvalue decomposition (scree plot) displaying percentage variance explained by each component, with elbow criterion supporting 2-component retention (cumulative 71.6% variance). (D) Variable correlation circle illustrating relationships between original strength measures and principal components, with color intensity representing contribution magnitude (variables near circle perimeter contribute most strongly to component structure).

4. Discussion

The present study demonstrated that isokinetic knee and hip extension strength are the strongest predictors of vertical jump performance in elite wrestlers, as evidenced by the robust correlation between knee extension torque and double-leg countermovement jump. The observed correlation coefficient of R = 0.798 between knee extension strength and bilateral jump performance represents one of the strongest relationships reported in athletic populations, notably exceeding those in team sports like soccer (R = 0.45–0.62) but similar to combat sports like boxing (R = 0.72–0.85),^{[17,21,28,29,43,44]} as well as different age groups.^[45] This highlights the sport-independent relevance of knee extensor strength as a general determinant of explosive lower-limb function even in non-athletic adult populations.^[9]

In literature, many studies investigating the relationship between isokinetic strength and jump performance have predominantly focused on the knee joint,^[28–30,46] due to its crucial role in force generation during the propulsive phase of vertical jumping.^[4,5,47] Moreover, the majority of these studies have evaluated bilateral jump tasks, such as the CMJ and squat jump, while research focusing on unilateral jumping remains relatively limited.^[30,46,48] Several studies have reported that this relationship was not consistently observed at lower angular velocities during isokinetic testing, particularly at or below 60°/s, where correlations with jump performance were weak or statistically insignificant.^[5,21,22]

The lack of association was explained by several factors, including the muscle force–velocity characteristics,^[5] the biomechanical differences in neuromuscular activation patterns between open-chain (isokinetic tests) and closed-chain tasks (jumping),^[49] and the contribution of elastic energy recoil mechanisms during the stretch-shortening cycle in jump execution.^[50] However, several previous studies have reported contrasting significant associations between isokinetic knee extensor strength and bilateral vertical jump performance, with indicated values ranging from R = 0.50 to 0.75, which is consistent with the present study and reinforces the pivotal role of the knee extensors in vertical jumping tasks.^{[19,29,51,52]} At the same time, elite boxers demonstrated similar correlations (R = 0.72–0.85) between lower-limb extension strength and explosive performance.^[28] The magnitude of associations in our wrestlers (R = 0.798) exceeds those reported in team sports like soccer (R = 0.45–0.62) or basketball (R = 0.51–0.68), potentially reflecting the sport-specific neuromuscular adaptations from wrestling’s emphasis on explosive hip and knee extension during techniques.^[29,46] The superior predictive capacity observed in our study (68.8% variance explained) compared to previous reports (38–42%) likely reflects our comprehensive multi-joint assessment approach and FFM normalization, which better captures the true neuromuscular determinants of explosive performance.^[47] In support of this, a study involving 150 adolescent wrestlers demonstrated that FFM-based torque values were more strongly correlated with isokinetic strength across all angular velocities compared to body mass.^[53] Similarly, in another study, jump height showed no association with absolute isokinetic knee extension torque at 60°/s; however, moderate and statistically significant correlations were observed when torque was normalized to body mass.^[48] The 60°/s angular velocity was selected in this study based on previous research demonstrating optimal correlations with explosive movements at moderate velocities that more closely approximate the force–velocity characteristics of vertical jumping compared to slower (≤30°/s) or faster (≥180°/s) velocities.^[19,29] This velocity represents a compromise between maximum force expression and movement specificity for wrestling-related explosive actions. In addition, the current study specifically aimed to investigate the relationship between joint-specific strength and power production. Considering the principles of the force–velocity relationship, concentric contractions at a low angular velocity (60°/s) were preferred for isokinetic strength testing, as they more closely reflect the mechanical demands of explosive athletic tasks. Additionally, compared to single-leg jumps, double-leg CMJs are typically performed with greater knee flexion angles, which increases the range of motion over which the knee extensors can generate force. This enhanced mechanical advantage may contribute to the stronger association observed between knee extension strength and jump performance in bilateral conditions.^[4,54]

Moreover, the inclusion of hip extension strength significantly improved the model’s explanatory power, accounting for 68.8% of the variance in double-leg CMJ performance. The multi-joint coordination patterns assessed reflect the neuromuscular demands of wrestling-specific movements, where coordinated hip and knee extension generate the force necessary for explosive techniques and defensive maneuvers. The principal component structure, with the primary extensor strength component explaining 52.6% of the variance and loading strongly on hip and knee extension variables (loadings > 0.70). The principal component structure provides empirical support for a hierarchical organization of lower-limb strength, where the primary extensor component represents the fundamental neuromuscular capacity underlying explosive vertical movements. This 2-factor model aligns with biomechanical evidence that hip and knee extensors contribute 72% to 84% of total jump power,^[5] while auxiliary muscles provide stabilization and fine motor control. The factor loadings (>0.70) for hip and knee extension on the primary component suggest these muscle groups function as a coordinated unit during explosive tasks, supporting targeted training approaches that emphasize multi-joint extensor development rather than isolated single-joint strengthening. This finding provides empirical support for the biomechanical importance of coordinated hip–knee extensor function during explosive vertical movements.^[54] It aligns with previous studies that developed regression models incorporating hip and knee extensor torques to predict vertical jump performance, with explained variances ranging from 38% to 75%.^[9,22,47] For instance, Śliwowski et al found that linear combinations of isokinetic torques explained 38% and 42% of the variance in CMJ and squat jump height, respectively.^[22] Similarly, Nishiumi et al found comparable results, while another study demonstrated that hip and knee extension torque together predicted up to 75% of jump performance variance when power was calculated as jump height × body mass.^[47] Notably, it was only 1 study that evaluated isokinetic strength across all 3 major lower-limb joints and, consistent with our findings, observed no significant association with ankle joint strength.^[9] From a physiological perspective, both isokinetic (slow-velocity) and vertical jump tests (fast-velocity) preferentially recruit type II (fast-twitch) muscle fibers, which generate high force outputs at elevated contraction velocities characteristic of explosive movements.^[55,56] Elite power athletes demonstrate greater type II fiber cross-sectional area and higher myosin heavy chain IIa composition compared to endurance athletes,^[57] directly supporting maximal force production during wrestling-specific actions such as takedowns and lifts.^[55,58] Although movement velocities differ (e.g., isokinetic testing at 60°/s vs plyometric jump exercises), among the neural adaptations associated with strength development, the increased recruitment of high-threshold motor units appears to be one of the most important adaptations that is not velocity and task specific. In a study by Behrens et al, plyometric training induced gains in isometric, concentric, and eccentric strength that were partly mediated by neural mechanisms. After training, maximum voluntary torques increased by 20, 24, and 27 N·m for isometric, concentric, and eccentric maximum voluntary contraction, respectively, compared to controls.^[59] This indicates that training enhances motor unit recruitment, particularly of high-threshold units, which contributes to force production across different contraction modes. In the context of wrestling, where both maximal strength training (slow-velocity, high-load efforts) and high-velocity training (explosive, power-oriented exercises) are implemented, this adaptation is of particular relevance. The increased participation of high-threshold motor units not only supports maximal force generation during heavy resistance exercise but also facilitates rapid force production during explosive actions, thereby enhancing overall power output. Such neuromuscular adaptations likely explain the transferability of strength and power gains to wrestling-specific movements that demand both force and speed. The 60°/s isokinetic testing protocol emphasizes maximal concentric force production at a moderate velocity, closely mimicking functional sport-specific movements. In contrast, the CMJ depends on the stretch-shortening cycle, utilizing stored elastic energy and reflexive motor unit recruitment to boost power output. Biomechanically, the hip and knee extensors showed strong contributions to the primary component in our PCA (loadings > 0.70), underscoring their crucial role in vertical propulsion. The gluteus maximus, hamstrings, and quadriceps work together to produce both isokinetic torque and explosive jump force. This coordination (especially among type II fiber-dominant proximal extensors) forms the physiological and biomechanical basis for the observed strength–performance relationships in elite wrestlers. Collectively, these findings explain why hip and knee extensors are more predictive of performance than isolated joint measures and support training approaches that focus on multi-joint, high-power movements.

The principal component structure provides empirical support for a hierarchical organization of lower-limb strength, where the primary extensor component represents the fundamental neuromuscular capacity underlying explosive vertical movements. This 2-factor model aligns with biomechanical evidence that hip and knee extensors contribute 72% to 84% of total jump power,^[5] while auxiliary muscles provide stabilization and fine motor control. The factor loadings (>0.70) for hip and knee extension on the primary component suggest these muscle groups function as a coordinated unit during explosive tasks, supporting targeted training approaches that emphasize multi-joint extensor development rather than isolated single-joint strengthening. Inter-limb asymmetry analysis revealed that 33.3% of wrestlers exceeded the 10% threshold for hip extension strength, indicating clinically relevant imbalances. This finding aligns with previous literature suggesting that asymmetries above 10% to 15% are indicative of neuromuscular inefficiency and increased injury susceptibility.^[60,61] While these asymmetries showed no association with bilateral jump performance, the moderate correlation between hip extension asymmetry and unilateral jump performance differences (R = 0.441, P = .042) reflects the task-specific nature of asymmetry effects, as unilateral explosive actions are more sensitive to strength imbalances. Recent meta-analyses and experimental research show that larger asymmetries in hip and hamstring strength are generally more detrimental to single-limb explosive tasks (like single-leg vertical jumps), and less predictive for bilateral tasks such as 2-leg vertical jumps.^[62,63] Our results reinforce the current consensus that practitioners should use both unilateral and bilateral assessments to capture functionally relevant asymmetries. Therefore, combat sport athletes, including wrestlers, have high demands for rapid, multidirectional single-limb actions, underlining the importance of asymmetry monitoring, inclusion of asymmetry screening in wrestler assessment protocols, particularly when values exceed the 10% cutoff, to identify suboptimal neuromuscular function or heightened injury risk.^[60] The reduced explanatory power in unilateral conditions (dominant leg: 39.8%, non-dominant leg: 25.7%) compared to bilateral performance (68.8%) indicates that bilateral assessments provide superior prediction accuracy for wrestler-specific explosive actions. This highlights the necessity of integrating both unilateral and bilateral measures into athlete monitoring to detect hidden deficits and guide corrective training.^[63] Specifically, hip and knee extension strength were significant predictors in the dominant leg model, together explaining 39.8% of the variance. In contrast, only hip extension remained a significant predictor in the non-dominant leg model (R² = 0.307). This asymmetry may be partly attributed to neuromuscular efficiency and habitual motor recruitment patterns developed through sport-specific dominance. Supporting this, 1 study observed a stronger relationship between isokinetic knee extensor torque and unilateral CMJ height in the dominant leg (R = 0.72) than in the non-dominant leg (R = 0.59), further highlighting limb-specific performance differences in explosive tasks.^[30] Additionally, lumbopelvic–hip complex stability is a critical factor in maximizing force output during unilateral stance, with instability potentially limiting peak performance on the non-dominant side.^[61] Given that a single-leg vertical jump requires overcoming the same body weight and inertia with only 1 limb, the neuromechanical demands are inherently elevated.^[48] Moreover, double-leg and single-leg hops display divergent force–velocity profiles,^[64] and supporting this, 1 study noted that unilateral vertical jumps significantly alter the force–velocity–power relationship by reducing the shortening velocity of leg extensors, thereby shifting the mechanical demand to larger proximal muscle groups.^[65,66] This biomechanical shift may explain the observed link between hip adduction–hip flexion strength and dominant leg jump performance, potentially due to the adductors’ role in hip extension when the hip is flexed and the hip flexors’ contribution to trunk stabilization during single-leg take-off.^[67] Collectively, these findings reinforce contemporary combat sport research emphasizing comprehensive asymmetry monitoring to inform training interventions aimed at reducing injury incidence and optimizing performance.

This study is distinguished by its comprehensive evaluation of all major lower-limb joints, the inclusion of both unilateral and bilateral jumping tasks, and the use of relative strength normalized to FFM. The identification of hip and knee extension strength as primary predictors has direct implications for wrestling-specific strength training, suggesting that targeted development of these muscle groups should form the foundation of power enhancement programs.^[68] Additionally, the involvement of elite athletes and the application of validated measurement tools further enhance the relevance and methodological rigor of the findings.

Despite its methodological strengths, this study has several limitations. The relatively small sample size may limit the generalizability of the findings and result in wide confidence intervals around specific regression coefficients. Although some variables were included in the regression models, they did not reach statistical significance when considered individually, possibly due to multicollinearity or limited statistical power. Additionally, the lack of electromyographic data prevents a direct understanding of neuromuscular activation strategies, and the cross-sectional design limits causal inference. Furthermore, inconsistent correlations across different angular velocities, especially at slower speeds, reflect previously reported variability and should be interpreted with caution. While maximum power output was prioritized as the primary performance outcome given its direct relevance to wrestling-specific explosive actions, additional variables including fatigue resistance, elastic efficiency, and alternative jump modalities (squat jump, drop jump) were excluded due to testing time constraints and participant fatigue management considerations. Comprehensive multi-joint isokinetic assessment across 8 bilateral movement patterns combined with 3 jump conditions already imposed substantial neuromuscular demands. Including additional fatigue-inducing protocols risked compromising data quality in primary outcome measures. Future investigations employing repeated-measures designs across multiple sessions could systematically evaluate these supplementary metrics alongside maximal strength and power capacities.

4.1. Practical applications

This study provides practical guidance for strength and conditioning professionals working with elite wrestlers and other combat sport athletes. Identifying hip and knee extension strength as key predictors of vertical jump performance emphasizes the importance of multi-joint, extensor-focused exercises like squats, lunges, and Romanian deadlifts. These exercises are beneficial for improving wrestling-specific explosive actions such as lifts, throws, and quick positional recoveries. Using FFM-normalized torque values enhance the accuracy of strength assessments, offering a more personalized evaluation tool for athletes in weight-class sports. Practitioners are encouraged to adopt this method for fairer comparisons and improved performance tracking. The finding that > 10% inter-limb asymmetry occurred in one-third of the sample highlights the need for regular asymmetry screening, especially for hip extension strength. Although asymmetries had little effect on bilateral jump performance, they were moderately linked to unilateral performance, underscoring their functional importance. Because unilateral tasks show reduced predictive power (particularly for the non-dominant limb) training and testing should include both bilateral and unilateral exercises to identify and correct side-to-side imbalances. Lastly, the PCA identified a primary extensor strength factor, supporting the use of combined, multi-joint strength assessments over isolated tests when creating athlete profiles and personalized performance programs.

5. Conclusions

This study demonstrated powerful predictive relationships between multi-joint isokinetic strength and vertical jump performance in elite wrestlers, with knee extension and hip extension strength explaining 68.8% of the variance in bilateral jumps through a primary extensor strength factor. PCA revealed a clear 2-factor structure of strength, with the primary extensor component representing the fundamental neuromuscular capacity underlying explosive lower-limb function. These findings establish multi-joint isokinetic assessment as superior to single-joint evaluation for predicting wrestling performance, with FFM normalization providing enhanced precision. These findings support the implementation of hip and knee extension-focused training protocols for improving explosive power in wrestling, with FFM-normalized strength assessments providing superior performance prediction compared to traditional approaches.

Author contributions

Conceptualization: Ozan Sever, Halil İbrahim Ceylan.

Data curation: Kübra Özdemir, Ferhad Jalalov, Serdar Bayrakdaroğlu.

Formal analysis: Wissem Dhahbi, Halil İbrahim Ceylan.

Funding acquisition: Valentina Stefanica.

Investigation: Hasan Hüseyin Yilmaz, Serdar Bayrakdaroğlu, Kübra Özdemir, Ahmet Gökhan Yazici, Ahmet Rahmi Günay.

Methodology: Ozan Sever, Halil İbrahim Ceylan.

Project administration: Halil İbrahim Ceylan.

Resources: Hasan Hüseyin Yilmaz, Kübra Özdemir, Ahmet Gökhan Yazici, Serdar Bayrakdaroğlu, Ahmet Rahmi Günay.

Software: Wissem Dhahbi, Halil İbrahim Ceylan.

Supervision: Halil İbrahim Ceylan.

Writing – original draft: Ozan Sever, Halil İbrahim Ceylan, Wissem Dhahbi, Valentina Stefanica.

Writing – review & editing: Ozan Sever, Halil İbrahim Ceylan, Hasan Hüseyin Yilmaz, Valentina Stefanica.