Secondary Logo

Journal Logo

Original Research

Biomechanical Comparison of Unilateral and Bilateral Power Snatch Lifts

Lauder, Mike A; Lake, Jason P

Author Information
Journal of Strength and Conditioning Research: May 2008 - Volume 22 - Issue 3 - p 653-660
doi: 10.1519/JSC.0b013e3181660c89
  • Free



Recently, there has been an increase in the inclusion of unilateral weightlifting movements into athlete strength and conditioning (S &C(JM1)) programs (7,17,18). Coaches have rationalized that the observed asymmetric nature of these lifts places different neuromuscular demands on their athletes: securing a heavy dumbbell overhead at the conclusion of a power snatch or jerk requires great strength in the trunk and shoulder girdle musculature (1,4,7,14,17,18). It is also believed that because these movements are less technically demanding, and thus easier to learn, they provide a simple but effective method of implementing resistance-training variety into athlete S&C programs (1,4,7,14,17,18).

Differences between the performance characteristics of unilateral and traditional bilateral weightlifting movements have not been quantified. The observed asymmetric nature of unilateral weightlifting movements suggests corresponding asymmetric force and movement patterns. Research into this area could aid the S&C coach with regard to the suitability of unilateral weightlifting movements within their athletes' S&C programs.

Biomechanically, variations of traditional bilateral barbell weightlifting movements have been extensively researched (3,8-10,12). Researchers have analyzed 2-dimensional footage of weightlifting movements in the sagittal plane (8,9,16), in addition to ground reaction forces (GRF) recorded from one force platform (8,21). Results have shown the typical patterns of force that are applied through the feet during weightlifting movements and the subsequent bar and body movements (8,16,19), which has provided important information for the coach of both competitive weightlifters and those using the lifts to enhance their athletic potential. In addition to this, research findings have highlighted mechanical similarities between variations of weightlifting movements and many sporting movements (6,11,15,23).

A limitation to 2-dimensional video and single force platform analysis is that it cannot identify movement asymmetries (7). Typically, this limitation does not present a problem because, by their nature, traditional bilateral weightlifting movements occur predominantly in the saggital plane. However, movement asymmetry has been reported during weightlifting performance (19). In addition to this, evidence exists to suggest that 2-dimensional video analysis may ignore bar and/or body rotation, suffer field of view obstructions, and inaccurately project body angles (3,10). Using synchronized 3-dimensional video and 2 force platform configurations, researchers at the U.S. Olympic training center have overcome these limitations, identifying asymmetrical movement patterns in weightlifters (19). This information is critical to athletic well being because it allows the coach to identify deviations from movement symmetry and manipulate lifting technique and/or training accordingly with a view to improving performance whilst minimizing injury potential (19). Conversely, the use of these analysis methods to examine the biomechanical characteristics of unilateral weightlifting movements and comparing them with the biomechanical characteristics of their traditional bilateral weightlifting movement variations would allow researchers to quantify the validity of anecdotal evidence of S&C coaches (7,17,18) regarding the asymmetrical nature of unilateral weightlifting movements.

With this in mind, the purpose of this study was to examine and compare the biomechanical characteristics of a unilateral and bilateral snatch movement. It was hypothesized that the applied force patterns and subsequent bar and body movements of the unilateral movement would deviate significantly both in terms of within lift symmetry and between lift characteristics. Support for this hypothesis would go some way to validate the claims made in the literature (1,4,7,14,17,18), providing potentially valuable information for S&C coaches.


Experimental Approach to the Problem

To examine whether there were any differences between selected biomechanical characteristics of a unilateral and bilateral weightlifting movement, subjects performed both a unilateral dumbbell (DBPS) and a traditional barbell power snatch (BBPS), in that order, with a resistance that corresponded with of 80% of their one repetition maximum (1 RM) (19) during one testing session. Force and video data of both the lifting (the side corresponding with the hand holding the DB) and nonlifting side were recorded from 3 trials of each movement with the use of high-speed 3-dimensional video techniques and 2 force platforms. Two-way repeated measures analysis of variance was used to test for significant differences between the dependent variables of bar linear displacement and velocity, hip, knee and ankle angular displacement and velocity, and vertical ground reaction force (VGRF) parameters, which included peak pull phase VGRF, catch phase VGRF, VGRF rate of change, and catch phase loading rate (LR), in addition to the independent variables of DBPS and BBPS and lifting and nonlifting side.


Following ethical clearance by the university and a thorough explanation of the aims and experimental procedures, 10 male weightlifters volunteered to participate in this investigation. Each volunteer had a minimum of 1 year's experience with both the DB and BBPS, regularly including them as part of their S&C training. All subjects provided written informed consent. The mean (±SD) subject physical and load characteristics are displayed in Table 1.

Table 1
Table 1:
Subject physical and bar load characteristics. (n = 10)


After a thorough warm up, participants were instructed to work up with progressively heavier single lifts to a one repetition maximum (1RM) for that day. The bar was then loaded to 80% of this 1RM value, and 3 single-lift trials were performed (19). This process was performed first with the DBPS and then the BBPS because, it was reasoned, the DBPS being the lighter lift would not impact on the performance of the heavier lift to an extent that was detrimental to the subsequent BBPS performance, whereas the opposite might not be true. These performances were filmed and synchronized ground reaction forces recorded, with volunteers resting as needed between lifts (19). The use of the 80% of 1RM testing load was based on the methodology of Reiser et al. (19), who cite this as a typical training load. During the BBPS, an Olympic weightlifting bar (Eleiko, Sweden) was loaded with IWF approved bumper plates. The bar used during the DBPS was a spin-lock DB bar loaded with 10kg weight plates that had an outer circumference of 28 cm. The DB and BBPS are sequentially illustrated in Figures 1 and 2, respectively.

Figure 1
Figure 1:
Photo sequence of DBPS performance during the present investigation.
Figure 2
Figure 2:
Photo sequence of typical BBPS during the present investigation.


Two gen-locked video cameras (PS200, Peak Performance Technologies Inc, Englewood, Colo) were positioned 5 m from the center of the force platforms with an inter camera angle of ∼90°. They were used to film the lifting trials at 200 Hz with a shutter speed of 1/1000 seconds (12) after first filming a 25-point 3-dimensional calibration frame (Peak Performance Technologies Inc.). This allowed the subsequent calculation of the 3-dimensional spatial coordinates of the bar ends, and lower limb joints using the direct linear transformation procedure. Successful trials for each subject representing 80% of 1RM on both lifts were digitized at 200 Hz using the Peak Motus 32 software, from the initial pull to the catch. After digitization, the raw coordinate data were smoothed using a low pass filter (12) and the kinematic parameters of interest taken directly from the program. An Opus technologies personal computer running Bioware 3.21 software (Kistler Instruments Ltd., Amherst, NY) recorded the VGRF of both feet at a sampling frequency of 500 Hz from 2 parallel 0.4 m by 0.6 m Kistler 9281 force platforms (Kistler Instruments Ltd.) that were positioned with the 0.6-m edges together. An external trigger mechanism synchronized the video and force platform data.


The angular kinematics of the lifting and nonlifting lower limb was calculated and lift-phases determined from changes in knee angular displacement (12,13). For the purpose of this study, the body side lifting the dumbbell was referred to as the lifting side, with the other side referred to as the nonlifting side. Bar kinematics were determined for between lift comparisons. The sequence and magnitude of the lifting and nonlifting side vertical ground reaction forces were determined, as were rate of force change during the pull and catch loading rates calculated for comparison. Thus, the independent variables were the kinetic and kinematic data, and the dependent variables the 2 different weightlifting movements and lifting and nonlifting side. Although both sides of the body were lifting during the BBPS the terms lifting and nonlifting were still used enabling the study of asymmetry during the BBPS, as well as between lift comparisons.

Statistical Analyses

Intraclass correlation coefficients (ICCs) were calculated for each of the kinetic and kinematic variables to indicate test-retest reliability. A 2-way repeated measures analysis of variance was used to determine between and within lift and within side differences, with paired t-tests (applying the Bonferonni correction) performed on significant results. The level of statistical significance was set at p ≤ 0.05.


Test-Retest Reliability

All kinetic and kinematic parameters demonstrated high test-retest reliability with ICC values consistently within the ICC range of 0.92-0.95.

Joint Kinematics

Typical hip, knee, and ankle angular displacements during the DBPS deviated from the patterns typically found in BBPS pulling movements. This is illustrated in Figure 3, which also shows the degree of movement symmetry between the lifting and nonlifting sides. The 2 lifts were similar in 3 areas: the start, maximum pull and catch, and the mean (±SD) joint kinematics of these phases are presented in Table 2. Statistical analysis showed no significant differences between lifting and nonlifting side angular displacements or angular velocities during the BBPS (p > 0.05, Table 2), but significant symmetrical deviations were demonstrated during the DBPS (p < 0.05, Table 2). Significant differences also were found between the BBPS and DBPS angular displacements and angular velocities (p ≤ 0.05, Table 2).

Table 2
Table 2:
Peak angular displacements and peak angular velocities of the hip, knee, and ankle for the lifting (L) and nonlifting (N) sides at the start (a), maximum extension of the knee (b), and catch (c) for the barbell (BB) and dumbbell (DB) lifts.
Figure 3
Figure 3:
Typical BBPS and DBPS hip, knee and ankle angular displacements. (a) indicates the start of the lift; (b) indicates the maximum extension of the knee, and c the catch.

Figure 3 illustrates the typical pattern of movement for the lifting and nonlifting sides at the hips, knees and ankles. The key moments of the start of the pull phase (i), the maximum extension of the knee (ii), and the catch (iii) are highlighted for both lifts.

Bar Kinematics

During the BBPS, a bar horizontal displacement pattern of towards, away, and towards the lifter was recorded. However, during the DBPS the bar was horizontally displaced away from and then towards the lifter. The typical bar trajectories, and mean (±SD) values, are graphically presented in Figure 4.

Figure 4
Figure 4:
Typical BBPS and DBPS bar trajectory, with horizontal displacement away from (positive) and toward (negative) the lifter.

The DB demonstrated significantly greater vertical displacement compared with the BB (p < 0.001, DBPS: 1.706 ± 0.095 m; BBPS: 1.417 ± 0.07 m), and catch towards horizontal (p < 0.05, DBPS: 0.27 ± 0.19 m; BBPS: 0.15 ± 0.05 m). In addition to this, peak DB vertical velocity was significantly greater than the peak BB vertical velocity (p < 0.001, DBPS: 3.17 m·s−1; BBPS: 2.18 m·s−1).

Ground Reaction Forces

Typical BBPS and DBPS VGRF patterns are displayed in Figures 5 and 6, with mean (±SD) VGRF data, along with symmetry and lift differences shown in Table 3.

Table 3
Table 3:
Vertical ground reaction forces: peaks (Newtons and body weights), rate of change (N.spen]1) and loading rate (BW.s−1).
Figure 5
Figure 5:
Typical BBSP VGRF patterns. The period between (a-b) indicates the first pull, (b-d) knee flexion, (d-e) the second pull, (e-f) unweighting, and (f-g) the catch loading phase.
Figure 6
Figure 6:
Typical DBPS VGRF. The period between (a-b) indicates lifting side knee extension and nonlifting side knee and hip extension, (b-c) lifting side hip extension, (c-d) unweighting, and (d-e) catch loading phase.


The power snatch requires a loaded bar to be lifted from the floor to overhead in one movement and is dependent upon the sequence and magnitude of the force applied to the bar and the subsequent lower limb angular displacements (2,13). The results of this investigation show that the type of bar used influences these factors.

Barbell snatch lift variations require the bar to be vertically displaced to and around the knees, before it continues to the overhead catch position. The BBPS pattern of vertical ground reaction forces (Figure 5) and the consequent lower limb angular displacements (Figure 3) and bar horizontal displacements (Figure 4) were consistent with the double knee bend pulling technique described in the literature (5,8,9,12,13,16). In addition, the assumption of symmetrical movement was also supported within this subject population.

The DBPS bar and body movement and VGRF data significantly deviated from the corresponding BBPS data. The DB was positioned between the lifters feet and the necessary trunk inclination towards the lifting leg to attain the start position required significantly greater lifting leg hip flexion (p < 0.02). The DB's position facilitated uninterrupted hip, knee, and ankle angular displacement to their maximum values. This displacement is graphically illustrated in Figure 3, which also illustrates the degree of asymmetrical hip, knee and ankle angular displacement. This is reflected by the VGRF and bar trajectory data shown in Figures 4 and 6.

The DBPS VGRF data suggest that the lifting leg angular displacements were consistent with double knee bend style of pull (8). However, the lower limb angular displacement data showed that the initial peaks were consistent with the peak lifting and nonlifting knee and nonlifting hip extension occurring within 0.047 seconds of one another. The delay in the peak lifting hip extension explains the nonlifting legs significantly greater first VGRF peak (p < 0.005) and the subsequent lifting leg VGRF peak, which was the result of a greater hip extension (p < 0.001) and angular velocity (p < 0.02). The greater BBPS lifting leg first and lifting and nonlifting leg second VGRF peak reflect the significantly greater loads used during the BBPS (p < 0.001).

There was no significant difference between the BBPS and DBPS catch VGRF (p > 0.05). However, Figure 6 illustrates that the DBN catch VGRF was significantly greater than that of the DBL (p < 0.03). This, combined with the 5° and 10.6° less DBN hip and knee flexion at the catch, which resulted in the DB nonlifting leg generating a significantly greater catch loading rate than the lifting leg (p < 0.05).

The combination of DB plate circumference and its frontal pull plane of motion caused a greater bar “loop” (p < 0.05). The “loop” describes the degree of bar horizontal displacement between its furthest forward and rearward catch position, and excessive “looping” has been found in novice lifters (24), and during failed lifts attempts (16,22). The DB demonstrated a significantly greater vertical displacement (p < 0.001) because of its central, single handgrip instead of the typical wide snatch grip (13,15), and the lower DB start position. The DB peak vertical velocity was significantly greater than that of the BB (p < 0.0003) because of the difference in the weight of the bar.

Practical Applications

The biomechanical study of weightlifting movements allows S&C coaches to gain insight about their biomechanical characteristics. This enables the coach to assess their relevance to, and safety in athlete S&C programs (19,21). This study found that the nonlifting side tended to generate a greater pull phase VGRF significantly faster (p = 0.001) than the lifting side during the DBPS. In addition to this, nonlifting side catch loading rates were almost double those of the lifting side. These results quantify the effects of a unilateral Olympic lift variation on movement patterns both during the concentric muscular contraction of load vertical displacement, and the loading implications of unilateral landing. The sporting implications of these results are important for S&C coaches as they support the suggestion of unilateral movements providing a different training stimulus (7,17,18), and their application to S&C for sports specific movements, including unilateral body movement and shock absorption (2,4,18,20). Further biomechanical study, including electromyographical analysis, would enhance these findings by determining the effect of this type of training on muscular recruitment.


1. Auferoth, SJ and Joseph, J. The dumbbell power clean. Natl Strength Cond Assoc J 12: 47-48, 1990.
2. Bartonietz, KE. Biomechanics of the snatch: Toward a higher training efficiency. Strength Cond 18: 24-31, 1996.
3. Baumann, W, Gross, V, Quade, K, Galbierz, P, and Schwirtz, A. The Snatch technique of world class weightlifters at the 1985 world championships. Int J Sport Biomechanics 4: 68-89, 1988.
4. Blackwood, B. Clean variations for enhanced performance. Strength Cond J 26: 74-76, 2004.
5. Burdett, RG. Biomechanics of the snatch technique of highly skilled and skilled weightlifters. Res Q Exerc Sport 53: 193-97, 1982.
6. Canavan, PK, Garrett, GG, and Armstrong, LE. Kinematic and kinetic relationships between an Olympic style lift and the vertical jump. J Strength Cond Res 10: 127-130, 1996 .
7. Cross, T. Rationale and coaching points for Olympic style lifting to enhance volleyball performance. Natl Strength Cond Assoc J 15: 59-61, 1993.
8. Enoka, RM. The pull in Olympic weightlifting. Med Sci Sports Exerc 11: 131-137, 1979.
9. Garhammer, J. Power production by Olympic weightlifters. Med SciSports Exerc 12: 54-60, 1980.
10. Garhammer, J. Weightlifting performance and techniques of men and women. In: Proceedings of First International Conference on Weightlifting and Strength Training (conference book). Lahti, Finland, 1998. pp. 89-94.
11. Garhammer, J and Gregor, R. Propulsion forces as a function of intensity for weightlifting and vertical jumping. J Appl Spor Sci Res 6: 129-134, 1992.
12. Gourgoulis, V, Aggelousis, N, Mavromatis, G, and Garas, A. Three-dimensional kinematic analysis of the snatch of elite Greek weightlifters. J Sport Sci 18: 643-652, 2000.
13. Hakkinen, K and Kauhanen, H. A biomechanical analysis of selected assistant exercises of weightlifting. JHumMovStud 12: 271-288, 1986.
14. Hedrick, A. Dumbbell training for football at the US Air Force Academy. Strength Cond 20: 34-39, 1998.
15. Hoffman, JR, Cooper, J, Wendell, M, and Kang, J. Comparison of Olympic vs traditional power lifting training programs in football players. J Strength Cond Res 18: 129-135, 2004.
16. Isaka, T, Okada, J, and Funato, K. Kinematic analysis of the barbell during the snatch movement of elite Asian weightlifters. J Appl Biomech 12: 508-516, 1996.
17. Marsit, J. Strength and conditioning for women's basketball. Strength Cond 16: 70-74, 1994.
18. Melanson, P. Power training using dumbbells. NSCA Performance Training J 1: 12-19, 2002.
19. Reiser, RF, Smith, SL, Rattan, R. Science and technology to enhance weightlifting performance: The Olympic program. Strength Cond 18:43-51, 1996.
20. Santana, J-C. Single-leg training for 2-legged sport: Efficacy of strength development in athletic performance. Strength Cond J 23:35-37, 2001.
21. Souza, AL, Shimada, SD, and Koontz, A. Ground reaction forces during the power clean. JStrength Cond Res 16:423-427, 2002.
22. Stone, MH, O'Bryant, HS, Williams, FE, and Johnson, RL. Analysis of bar paths during the snatch in elite male weightlifters. Strength Cond 20: 30-38, 1998.
23. Stone, MH, Sanborn, K, O'Bryant, HS, Hartman, M, Stone, ME, Proux, C, Ward, B, and Hruby, J. Maximum strength and power-performance relationships in collegiate throwers. J Strength Cond Res 17: 739-745, 2003.
24. Winchester, JB, Erickson, TM, Blaak, JB, and McBride, JM. Changes in bar path kinematics and kinetics after power clean training. JStrength Cond Res 19: 177-183, 2005.

force-time curve; rate of force development; ground reaction force; kinematics; resistance training.

© 2008 National Strength and Conditioning Association