Secondary Logo

Journal Logo

Original Research

Lower Extremity Biomechanics During Weightlifting Exercise Vary Across Joint and Load

Kipp, Kristof1; Harris, Chad2; Sabick, Michelle B3

Author Information
Journal of Strength and Conditioning Research: May 2011 - Volume 25 - Issue 5 - p 1229-1234
doi: 10.1519/JSC.0b013e3181da780b
  • Free



Adaptations to resistance training programs are stimulus specific (10,15). Out of the number of program design variables that determine the extent of these adaptations (e.g., sets and repetitions), the most salient factor is the magnitude of the external load (10). Variations in the magnitude of training loads elicit force- and velocity-specific adaptations (16,19). Load-specific adaptations are perhaps best illustrated in training for maximal muscle power output (16). It is well documented that training at optimal loads-those that maximize power-are most effective in improving maximal muscle power (16,23). The use of optimal loads during a training session is therefore highly important when specific adaptations, such as increased muscular power or strength, are a primary goal.

Although numerous training modalities are currently used to improve dynamic power or strength performance, resistance training programs that incorporate weightlifting exercises, or derivatives of these exercises, are known to elicit superior adaptations (4,15,22). Incorporating weightlifting methods produces greater and broader improvements in jumping and sprinting performance than traditional heavy resistance training exercise (22). Experienced weightlifters also exhibit greater fast-twitch fiber activation and more optimal timing when producing peak force and rate of force development (9). The efficacy of weightlifting exercises are thought to arise from a high degree of specicifity in that they are biomechanically similar to many explosive sports movements (4).

Previous research indicates that changes in the external load used with weightlifting exercises directly affect the biomechanical characteristics of these exercises (5,17). In general, it appears that performance-associated biomechanical characteristics are maximized at submaximal loads (5,16). Commonly analyzed variables are related to the trajectories or external kinetics associated with the barbell itself or the lifter-barbell system (5,11-13,17,18,24,25). For example, the magnitudes of ground reaction forces and power associated with the movement of the barbell-lifter system differ significantly between low and high loads (17). Further, the velocity of the barbell and its trajectories are also affected by changes in the external load (24). Although these variables provide important global information about the mechanics at the location of external constraints (i.e., the bar and ground), they do not provide information about the internal joint kinetics. Because specificity of training is a function of the task-inherent biomechanics, not simply the external movement characteristics (20), knowledge of internal joint kinetics would provide important descriptive information to be used in the program design process.

Lower extremity joint kinetics vary based on the external load (3,8). Enoka (8) showed that absolute magnitudes of joint power production differ to accommodate changes in external loads. Although this study also examined lower extremity joint angular velocities and net joint torques, these variables were not included in the analysis and not reported. It thus remains to be seen how changes in the external load may affect lower extremity kinematics and kinetics. In addition to joint velocities, net torques, and power, research also suggests that the ability to rapidly generate torques (i.e., rate of torque development [RTD]) may be a particularly important variable related to functional performance (1). Collectively, an understanding of how these variables change across loads at each of the lower extremity joints would facilitate the design of specific resistance programs that incorporate weightlifting exercises. However, surprisingly little is known about the load-dependent biomechanics of the lower extremity during weightlifting exercise. Therefore, the purpose of this study was to determine the effect of changing external loads on hip, knee, and ankle joint biomechanics during the pull phase of the clean.


Experimental Approach to the Problem

We hypothesized that the external load lifted during the pull portion of the clean exercise significantly influences the biomechanical demands of the lower extremity. The rationale for this investigation was that a more precise understanding of these biomechanical demands at each lower extremity joint will facilitate proper design of resistance programs that incorporate weightlifting exercises. To determine the effect of load on lower extremity biomechanics during the pull phase of the clean, we measured kinematic and kinetic data of the hip, knee, and ankle joints while subjects performed sets of cleans at 65, 75, and 85% of their respective 1 repetition maximum (RM). The percentages were chosen because they span a percentage range that has previously been shown to maximize the external kinematics and kinetics of the barbell or barbell-lifter system (5,17). The dependent biomechanical variables were joint angular velocities, net torques, power, and RTD and were chosen based on their mechanical relationship to lower extremity performance.


A basic power calculation indicated that to detect moderate within-group and between-group differences with statistical power of at least 0.80 at α < 0.05, a minimum of 10 subjects per group would be required. We thus recruited 10 subjects (9 men and 1 woman) to participate in this study. All subjects indicated that they had participated in a training program that involved weightlifting exercises for at least the previous 6 months. Six subjects were Olympic weightlifters, and 3 were collegiate throwers. All subjects participated in this study during an “off”-week in their preseason. One-repetition maximums were self-reported and current within 2 weeks of testing (Table 1). All subjects were deemed technically competent and representative of collegiate-level weightlifters by a national USA Weightlifting coach. Subject characteristics are presented in Table 1. All subjects provided written informed consent after reading an informed consent document approved by the University's Institutional Review Board.

Table 1
Table 1:
Subject characteristics.*


Data Collection Procedures

Before commencement of data collection, all subjects completed a warm-up that included lifting light loads between 35 and 50% of their estimated 1RM for the clean exercise. After the warm-up, subjects performed 2-3 repetitions each at 65, 75, and 85% of 1RM. All subjects were given 2-3 minutes of rest between each set. Kinematic and kinetic data were collected during each of the 3 sets.

Kinematic and Kinetic Data Analysis

During the data collection session, subjects performed the clean exercise while standing on 2 force platforms that were built into an 8 × 8-ft weightlifting platform. The force plates were mounted flush with the top of the platform. During the execution of each set of lifts, the positions of reflective markers attached to bony landmarks of the subjects' body were recorded with a 6-camera motion capture system (Vicon 612; Vicon Peak, Lake Forest, CA, USA). Kinematic data were collected at 250 Hz and filtered at 6 Hz. Kinetic data were collected at 1,250 Hz from the 2 force plates (Kistler) and filtered at 25 Hz. To establish a neutral anatomical position, a single static calibration trial was performed. Kinematic and kinetic data were exported and processed with custom software in MATLAB. The lower extremity was modeled as a 3-dimensional system of rigid links. Euler angle rotation sequences of flexion, abduction, and internal rotation of the distal segment were used to calculate ankle, knee, and hip joint angles. Angles were numerically differentiated with a central difference technique to obtain joint angular velocities. Body segment masses, center of mass locations, and mass moments of inertia were calculated from measured anthropometrics and published sex-dependent relationships (7). Kinematic and kinetic data then were combined and used to solve for ankle, knee, and hip joint torques with a conventional inverse dynamics approach in 3 planes of motion. Calculated joint torques represent net internal torques and thus reflect the net influence of all anatomical structures crossing a joint. Joint powers were calculated as the products of velocity and torque. Net joint torques were numerically differentiated to calculate rates of torque development. Only positive peak joint kinematic and kinetic variables were extracted for analysis and represent peak angular extension velocity, extensor torque, extensor RTD, and extensor power generation. All peak variables were averaged between the right and left legs and submitted to statistical analysis. Although the analysis used 3-D joint segment models and yielded variables in all 3 planes of motion, only sagittal-plane variables were analyzed because the pull phase of the clean primarily involves muscles in the said plane. Pilot testing indicated that kinematic and kinetic variables exhibit acceptable reliability (intracorrelation coefficient >0.90).

Statistical Analyses

Peak positive kinematic and kinetic variables from 3 sets of cleans were analyzed: 65, 75, and 85% of 1RM. Dependent kinematic and kinetic variables chosen for analysis were peak angular velocities, net torques, power, and RTD for the ankle, knee, and hip joint. Separate general linear analysis of variance models were used to test for differences in dependent variables. Each model consisted of a 3 × 3 (load × joint) analysis to test for within-subject differences (load) and for between-subject (joint) differences. Within-subject differences (i.e., across load) were treated as repeated measures. Assumptions of the test statistic were verified with Mauchly's test of Sphericity. Greenhouse-Geisser (GG) corrections were made when assumptions of sphericity were violated. Partial eta-squared (η2) and power values were used to help interpret the magnitude of main and interaction effects. In the absence of a significant interaction effect, data were pooled across load to compare differences between joints. Post hoc analysis consisted of paired and independent t-tests for comparisons among between and within-subject differences, respectively. The standard of proof to show statistical significance for all analyses was set at a level of α ≤ 0.05. All statistical analyses were performed using SPSS version 17.0 (SPSS, Chicago, IL, USA).


The effects of load on lower extremity joint velocities did not depend on the respective load (interaction p = 0.11, η2 = 0.184, power = 0.549; Table 2). However, joint velocities were significantly influenced by joint independently (main effect p < 0.05). Joint velocities were significantly larger for the knee and hip than for the ankle.

Table 2
Table 2:
Mean ± SD lower extremity joint angular velocities (°·s−1), net torques (N·m), powers (W), RTD (N·m·s−1) for the hip, knee, and ankle joint at loads of 65, 75, and 85% of 1RM.*

Lower extremity net joint torques depended on combined effects of load and joint (interaction p = 0.001, η2 = 0.384, power = 0.961; Table 2). Specifically, hip joint torque at 65% of 1RM was significantly smaller than at 75 and 85% of 1RM. Knee joint torque at 85% of 1RM was significantly smaller than at 75% of 1RM. Ankle joint torque at 85% of 1RM was significantly larger than at 65 and 75% of 1RM. Further, knee joint torque was significantly lower than hip torque at all loads but differed from ankle joint torque only at 65 and 85% of 1RM.

Lower extremity joint powers depended on combined effects of load and joint (interaction p = 0.024 − GG correction, η2 = 0.311, power = 0.848; Table 2). Specifically, knee joint power was significantly higher at 75% of 1RM than at 65 and 85% of 1RM. Ankle joint power at 85% of 1RM was significantly higher than at 65% of 1RM. Hip joint power did not vary with load. Lower extremity power did not vary across joints.

The effects of load on lower extremity joint RTD depended on the respective joint (interaction p = 0.014, η2 = 0.254, power = 0.829; Table 2). Hip joint RTD was significantly smaller at 65% of 1RM than at 75 and 85% of 1RM. Knee joint RTD was significantly larger at 75% of 1RM than at 85% of 1RM. Ankle joint RTD was significantly smaller at 85% of 1RM than at 75% of 1RM. Further, joint RTD did not vary across joints.


The external load lifted during the pull phase of the clean has a direct influence on biomechanics of the lower extremity. Generally, load effects appeared more evident in lower extremity kinetics than kinematics. Although joint angular velocities did not change across the load spectrum, net joint torque, power, and RTD did vary with the load lifted but in part depended on the respective joint in question. Joint velocities did, however, vary across joints in that angular velocities were higher at the hip and knee than at the ankle.

Increases in external loads generally resulted in greater task demands imposed on the lower extremity. Ankle joint torque was significantly greater at 85% of 1RM than at 65 and 75%, which may underscore the importance of forceful plantar flexion during the final pull phase of the clean as the external load increases. Hip joint torque increased from 65 to 75% of 1RM but appeared to plateau thereafter. Contrary to our observation that hip joint torque stabilizes once the load exceeds 75% of 1RM, Baumann et al. (3) observed that hip joint torque during competitive weightlifting attempts increased as barbell load increased. In combination, the load-associated increase in hip and ankle torque from 65 to 75 and 85% of 1RM compare well with studies that demonstrate higher ground reaction forces in response to elevated loads (17). Knee joint torque, however, decreased when the load was increased from 75 to 85% of 1RM. Although, mechanically the magnitudes of the ground reaction forces are a direct reflection of the summed total of the body's net joint torques, an increase in external load may thus not always increase the functional torque or strength demands imposed on a joint. Accordingly, it has been suggested that it is not necessarily the magnitude of joint torque produced by the knee extensors during weightlifting movements that is important, but rather the control of the moment arm of the ground reaction force with respect to the knee joint center (3). However, collectively these results indicate that to maximize net joint torques of the lower extremities and increase the demand imposed on the involved musculature, loads should generally exceed 75 and 85% for the hip and ankle joint, respectively. Maximizing knee joint torque on the other hand may be achieved with loads <85% of 1RM but may also involve more complex control.

Lower extremity joint torque-time curve characteristics, as assessed through rate of joint torque development, were also influenced by changes in external loads. This finding is contrary to previous research where load had little influence on the rates of ground reaction force development (17). This discrepancy illustrates the importance to consider internal kinetics when evaluating task-inherent biomechanics of weightlifting exercises. More specifically, the results showed that ankle joint RTD was significantly greater at 85% of 1RM than at 75%, whereas knee joint RTD was greater at 75% of 1RM than at 85%. Further, hip joint RTD increased linearly and reached a maximal point at 85% of 1RM. It is important to note that these results generally match those for the above-reported joint torques. For example, knee joint torque and RTD were both greater at 75% of 1RM than at 85%. Similarly, both ankle joint torque and RTD were greater at 85% of 1RM than at 75%, whereas hip joint torque and RTD were greater at 75% of 1RM than at 65%. Although we did not measure movement time, the total time of the pull phases of weightlifting movements does not change significantly as load increases (14). It thus appears that to achieve higher net torques during a movement of a constant time interval, it becomes necessary to increase the rate at which the torque is developed. Consequently, in much the same way that joint torques were maximized at specified loads, the rate at which at these torques were developed followed a similar load-dependent pattern.

In addition to observing load-dependent behavior for joint torque and RTD, the power generated at the ankle and knee joint was also significantly influenced by load. Ankle joint power was higher at 85% of 1RM than at 65% of 1RM, whereas power at the knee joint was maximal at 75%. These results compare well to previous findings that show power output associated with either the barbell or barbell-lifter system is maximized between 70 and 80% (5,17). Lower extremity joint power, as measured in this study, is calculated as the product of joint torque and joint angular velocity. However, joint angular velocities did not vary across the load ranges used in this study. Indeed, it appears that joint velocities are less subject to change as resistance is increased (14). Because joint velocities remained constant, the observed differences in power at the ankle and knee joint may consequently be the result of higher joint torques. Although we did not extract the joint angular velocity and joint torque at the instant of maximal power, qualitative analyses of the time histories indicated that maxima for knee and ankle torques, RTD and power all occurred toward the end of the pull-movement (i.e., the second pull). External power outputs, derived from barbell kinematics, are also highest during this phase (11,12). It appears that the load-dependent changes in maximal lower extremity biomechanics result from an interplay between the kinetic variables and display temporal patterns similar to those observed externally from barbell mechanics.

Apart from the load-dependent changes in lower extremity biomechanics during the pull phase, some differences were joint dependent. Most notably, joint angular velocity was smaller at the ankle than at the knee and hip. The hip and knee joint undergo greater angular excursions than the ankle during the pull phase of the clean. A greater range of motion may necessitate greater velocities, especially if the time of the movement remains constant. Joint torques, on the other hand, were generally larger at the hip and ankle than at the knee. The difference between knee joint torques compared to those of the other joints can be interpreted similarly to the earlier described mechanism that accounts for the load-dependent change in knee joint torque in that the control of the moment arm of the ground reaction force about the knee joint is a greater determinant of joint torque than the external load. Alternatively, the torques reported in the represent study represent only the net internal torques and do not account for the possibility of antagonistic coactivation. An increase in hamstring coactivation with an increase in external load may effectively balance an increase in knee extensor torque. In light of the important technical implications associated with the control of joint torques during movement, it may be prudent for future weightlifting research to investigate moment arm control and antagonistic coactivation with respect to lower extremity joint function during weightlifting exercises.

Although this study provides novel biomechanical insights, the results should be interpreted with caution. First, the subjects in this study were all of a similar experience level. An individual's training status can significantly influence load-dependent expression of muscular performance (2,6,21). Extrapolating and applying these results to either more or less trained individuals may result in erroneous exercise prescription and should thus be done with caution. Second, the results from this study provide only a cross-sectional perspective of load-associated changes in lower extremity mechanics. Because resistance training or feedback-based programs may influence load-dependent biomechanical characteristics across the loading spectrum (24,25), the need for continuous assessment of lower extremity biomechanics during weightlifting exercise is warranted. These limitations indicate that lower extremity kinetics should be constantly monitored and suggest a need for longitudinal studies. Nonetheless, the results provide an important first step to a better understanding of the biomechanical demands of weightlifting exercise and should prove useful in the program design process.

The results of this investigation suggest that the external load significantly affects lower extremity kinetics, whereas differences in lower extremity kinematics appear to be more joint dependent. Together, the examination of joint torque, RTD, and power indicated that lower extremity kinetics follow a load-dependent pattern. In general, the mechanical behavior at a joint was maximized through high rates of torque development that allowed for the generation of high joint torques. In the absence of load-associated changes in joint velocities, the increased joint torques served to maximize joint power. This pattern was most apparent at the knee and ankle joint, where joint kinetics were maximized at 75 and 85% of 1RM, respectively.

Practical Applications

The rationale for this study was that facilitating a better understanding of the biomechanical demands at each lower extremity joints would assist proper design of resistance programs that incorporate weightlifting exercises. Selecting and training at external loads that maximize either joint torque and power would be expected to result in superior strength and power performance, respectively. If the goal of a weightlifting-based resistance training program is to maximize joint torque, RTD, or power of the knee extensors and the ankle plantar flexors in relatively well-trained individuals, loads of 75 and 85% of 1RM should be chosen to target the knee extensors and ankle plantar flexors, respectively. If hip extensor torque and RTD are program goals, loads >75% of 1RM should be targeted, respectively. Training status and adaptations should also be considered and closely monitored when applying the findings of this study.


We would like to thank Josh Redden and Seth Kuhlman for assisting with data collection and processing.


1. Aagaard, P, Simonsen, EB, Andersen, JL, Magnusson, P, and Dyhre-Poulsen, P. Increased rate of force development and neural drive of human skeletal muscle following resistance training. J Appl Physiol 93: 1318-1326, 2002.
2. Baker, D. A series of studies on the training of high-intensity muscle power in rugby league football players. J Strength Cond Res 15: 198-209, 2001.
3. Baumann, W, Gross, V, Quade, K, Galbierz, P, and Shwirtz, A. The snatch technique of world class weightlifters at the 1985 world championships. Int J Sport Biomech 4: 68-89, 1988.
4. Chiu, LZ and Schilling, B. A primer on weightlifting: From sport to sports training. Strength Cond J 27: 42-48, 2005.
5. Cormie, P, McCaulley, GO, Triplett, NT, and McBride, JM. Optimal loading for maximal power output during lower-body resistance exercises. Med Sci Sports Exerc 39: 340-349, 2007.
6. Cronin, J and Sleivert, G. Challenges in understanding the influence of maximal power training on improving athletic performance. Sports Med 35: 213-234, 2005.
7. de Leva, P. Adjustments to zatsiorsky-seluyanov's segment interia parameters. J Biomech 29: 1223-1330, 1996.
8. Enoka, RM. Load- and skill-related changes in segmental contributions to a weightlifting movement. Med Sci Sports Exerc 20: 178-187, 1988.
9. Felici, F, Rosponi, A, Sbriccoli, P, Filligoi, GC, Fattorini, L, and Marchetti, M. Linear and non-linear analysis of surface electromyograms in weightlifters. Eur J Appl Physiol 84: 337-342, 2001.
10. Fry, AC. The role of resistance exercise intensity on muscle fibre adaptations. Sports Med 34: 663-679, 2004.
11. Garhammer, J. Performance evaluation of Olympic weightlifters. Med Sci Sports 11: 284-287, 1979.
12. Garhammer, J. Power production by Olympic weightlifters. Med Sci Sports Exerc 12: 54-60, 1980.
13. Haff, GG, Jackson, JR, Kawamori, N, Carlock, JM, Hartman, MJ, Kilgore, JL, Morris, RT, Ramsey, MW, Sands, WA, and Stone, MH. Force-time curve characteristics and hormonal alterations during an eleven-week training period in elite women weightlifters. J Strength Cond Res 22: 433-446, 2008.
14. Häkkinen, K, Kauhanen, H, and Komi, PV. Biomechanical changes in the Olympic weightlifting technique of the snatch and clean & jerk from submaximal to maximal loads. Scand J Sports Sci 6: 57-66, 1984.
15. Izquierdo, M, Häkkinen, K, Gonzalez-Badillo, JJ, Ibanez, J, and Gorostiaga, EM. Effects of long-term training specificity on maximal strength and power of the upper and lower extremities in athletes from different sports. Eur J Appl Physiol 87: 264-271, 2002.
16. Kaneko, M, Fuchimoto, T, Toji, H, and Suei, K. Training effect of different loads on the force-velocity relationship and mechanical power output in human muscle. Scand J Sports Sci 5: 50-55, 1983.
17. Kawamori, N, Crum, AJ, Blumert, PA, Kulik, JR, Childers, JT, Wood, JA, Stone, MH, and Haff, GG. Influence of different relative intensities on power output during the hang power clean: identification of the optimal load. J Strength Cond Res 19: 698-708, 2005.
18. Kawamori, N, Rossi, SJ, Justice, BD, Haff, EE, Pistilli, EE, O'Bryant, HS, Stone, MH, and Haff, GG. Peak force and rate of force development during isometric and dynamic mid-thigh clean pulls performed at various intensities. J Strength Cond Res 20: 483-491, 2006.
19. Moss, BM, Refsnes, PE, Abildgaard, A, Nicolaysen, K, and Jensen, J. Effects of maximal effort strength training with different loads on dynamic strength, cross-sectional area, load-power and load-velocity relationships. Eur J Appl Physiol Occup Physiol 75: 193-199, 1997.
20. Siff, MC. Supertraining. Denver, CO: Supertraining Institute, 2004.
21. Stone, MH, O'Bryant, HS, McCoy, L, Coglianese, R, Lehmkuhl, M, and Schilling, B. Power and maximum strength relationships during performance of dynamic and static weighted jumps. J Strength Cond Res 17: 140-147, 2003.
22. Tricoli, V, Lamas, L, Carnevale, R, and Ugrinowitsch, C. Short-term effects on lower-body functional power development: Weightlifting vs. vertical jump training programs. J Strength Cond Res 19: 433-437, 2005.
23. Wilson, GJ, Newton, RU, Murphy, AJ, and Humphries, BJ. The optimal training load for the development of dynamic athletic performance. Med Sci Sports Exerc 25: 1279-1286, 1993.
24. Winchester, JB, Erickson, TM, Blaak, JB, and McBride, JM. Changes in bar-path kinematics and kinetics after power-clean training. J Strength Cond Res 19: 177-183, 2005.
25. Winchester, JB, Porter, JM, and McBride, JM. Changes in bar path kinematics and kinetics through use of summary feedback in power snatch training. J Strength Cond Res 23: 444-454, 2009.

clean; power; rate of torque development

© 2011 National Strength and Conditioning Association