Power is one of the main determinants for athletic performances such as sprinting and jumping (2,4,11). It has been reported that performing exercises that produce high power greatly increased muscular power and, therefore, enhanced athletic performance (2,4,7,11). Weightlifting and its derivatives are considered highly effective training methods for power development because they produce among the greatest power during the second pull phase where maximal extension occurs at the hip, knee, and ankle, relative to other traditional resistance exercises (2–4,6,12,14). One way to classify weightlifting exercises is by the presence or absence of the phase of catching the bar: catching derivatives that include the catching phase (e.g., hang power clean [HPC]) and pulling derivatives that omit the catching phase (e.g., jump shrug, hang high pull [HHP]) (16). Several comparative studies have investigated the effect of different external loads on power output during weightlifting exercises. These studies reported that both the jump shrug and HHP generate greater power than the HPC, suggesting the superior benefits of pulling derivatives for improving power output (8,9,19,20).
In catching derivatives such as the HPC, there is a limitation regarding the bar height lifted because of the catching movements. If lifters perform the second pull with maximal efforts at lighter loads, the bar will be lifted above the height limit, leading to an incorrect execution of catching the bar; thus, lifters must voluntarily limit the peak bar height lifted. Because of the strong positive correlation between peak bar height and maximal bar velocity at the end of the second pull phase in the HPC, lifters cannot produce their maximal velocity at lighter loads, resulting in markedly decreased power output (21). Conversely, the height limit for pulling derivatives such as the HHP is relatively greater because of the absence of the catching movement; thus, the second pull phase can be performed in a more powerful manner. For these reasons, it is logical that catching derivatives produce less power than the pulling derivatives at lighter loads.
However, because the movements up to the second pull phase are identical between the catching and pulling derivatives, both exercises should generate comparable power when using sufficiently heavy loads requiring lifters to perform the second pull with their maximal efforts. From a mechanical perspective, the results from previous studies reporting that catching derivatives produce less power than pulling derivatives across the entire loading spectrum may not have accurately assessed the pure power characteristics of the exercises in question (8,9,19,20). There are two possible explanations for the results: loading range and study subjects. First, the loading range investigated in these studies (30–80% of 1 repetition maximum [1RM] of the HPC) might not have been sufficiently wide. Because it is recommended to use various loads in power training (4,7,11) and loads heavier than 80% 1RM are often used for weightlifting training programs (13), it is necessary to compare peak power between exercises using a wider range of loads including loads greater than 80% 1RM. Another possible factor for the difference is that the movements might not have been executed correctly, especially at heavier loads (18). The subjects included in previous studies were not weightlifters. Because it is not easy to perform ballistic movements such as weightlifting exercises with correct technique at near-maximum loads (18), subjects should be proficient at movements to compare the pure power characteristics.
Therefore, the current study compared power output between the HPC as a catching derivative and the HHP as a pulling derivative across a wide range of loads among weightlifters. We hypothesized that the HHP would produce greater power at light to medium loads, whereas there would be no significant difference at heavy to maximal loads. A portion of the data (power, force, and velocity during the HPC) included in this manuscript has been reported elsewhere (20).
Experimental Approach to the Problem
A within-subject repeated-measures design was used to compare the power, force, velocity, and bar height produced during the HPC and HHP at different external loads. The exercises (HPC and HHP) and loads (40, 60, 70, 80, 90, 95, and 100% 1RM of the HPC) were selected as independent variables. The jump shrug is a frequently used pulling derivative, and it has been used in several studies (8,9,17–20). The exercise movement resembles a jump squat in which lifters intend to jump as high as possible (20). Although these exercises are considered highly effective for power development, they may cause injuries because of weighted impact at landing, particularly with heavy loads (6). The landing force during the jump shrug was reported to be approximately double compared with that of the HPC and HHP (17). Considering the risk of injury, the jump shrug was not included in this study using maximal loads. The loading range was 40–100% 1RM of the HPC, and loads were gradually increased from lighter to heavier loads for injury prevention and correct execution of the movements. Previous studies (19,20) reported that the HHP produced markedly greater power outputs than the HPC at 45% 1RM, and thus, 50% 1RM was not used to prevent unnecessary fatigue. Peak power and the contributing factors for determining peak power (force at peak power [Fpp], velocity at peak power [Vpp], and peak bar height) were used as dependent variables. Peak power was considered as the central outcome, whereas Fpp, Vpp, and peak bar height were used as suboutcomes.
Eight competitive weightlifters with at least 5 years of experience in weightlifting between the ages of 19 to 27 participated in this study, mean ± SD; age, 21 ± 3 years; height, 169.0 ± 4.2 cm; body mass, 80.3 ± 14.9 kg; 1RM HPC, 125.6 ± 14.5 kg; and 1RM HPC, 1.59 ± 0.17 kg/body mass. All subjects have competed in national or international competitions. This study was approved by the Ethics Review Committee on Human Research of Waseda University, and each subject provided written informed consent. The subjects frequently perform the HPC and HHP in their training, and they were proficient in the movements. All sessions were conducted shortly after the preparatory phase in their yearly schedule. They participated in 3 sessions over 3 days: the first day for 1RM testing and the second and third days for the power testing. Each testing session was separated by 2–10 days.
Hang Power Clean 1 Repetition Maximum Testing
On day 1, 1RM testing was conducted. The HPC was performed as previously described (20). Briefly, (a) subjects stood still with the bar held at their upper thigh, (b) lowered the bar to the upper part of their knee, (c) drastically lifted the bar upward with countermovement, and (d) received the bar on their shoulders and collar bones in a quarter squat position. The subjects performed a warm-up of 2 sets of 5 repetitions using a 20-kg weightlifting bar. They then gradually increased the loads to 50, 70, 80, and 90% of their self-reported 1RMs and added 2–5 kg depending on their previous lifts until they reach their 1RMs. The session was terminated if subjects failed attempts 2 times in a row. The success criterion was to receive the bar with the subjects' upper thigh above parallel to the ground.
Power testing sessions were conducted on days 2 and 3. Each subject was randomly assigned to perform the HPC or HHP on one of the days. The HPC was performed as described for the 1RM testing. The HHP was performed as previously described as follows (20): from the starting position, subjects lowered the bar to a position just above the knees, and then pulled the bar upward by fully extending their ankles, knees, and hips with countermovement. The loads used in the power testing were determined by each subject's 1RM obtained during 1RM testing. As described for 1RM testing, subjects started the session with warm-up sets of either the HPC or HHP. They then performed the HPC or HHP on a force platform (0625, ACP, AccuPower; AMTI, Watertown, MA) at 40, 60, 70, or 80% 1RM with 2 attempts per load and at 90, 95, or 100% 1RM with one attempt per load. The rest periods were 2 minutes between attempts at 40, 60, 70, and 80% 1RM and 3 minutes between attempts at 90, 95, and 100% 1RM. Subjects were instructed to perform each attempt with their maximal effort. They consistently used hook grips during 1RM and power testing because the grip types have a significant influence on bar velocity, force, and power output (12). In addition, knee sleeves, tapes, and weight belts were used if needed, which were controlled between the sessions for each subject.
Ground reaction forces were obtained from the force platform at a sampling rate of 1,000 Hz. System (lifter's body mass + external load) acceleration and velocity were calculated from the force data using forward dynamics. System power was calculated as the force multiplied by the system velocity at each time interval. Peak power was produced during the second pull phases, at which time Fpp and Vpp were measured. In addition, for 2-dimensional video analysis, all attempts were recorded from a sagittal plane with a digital camera at 120 Hz (15,373,667; FLIR Integrated Imaging Solutions, Inc., BC, Canada). A reflective marker placed on the end of the bar was manually plotted from video footage. The pixel coordinates of digitized points were scaled to obtain a vertical position of the bar from the surface of the force platform. The highest position obtained during the exercise was considered as peak bar height. For lifts performed at 40, 60, 70, or 80% 1RM, the repetition that generated the greater peak power was used for comparison.
All data were expressed as the mean (SD). Normality was verified using the Shapiro-Wilk test. The intraclass correlation coefficient (ICC) was calculated to report the test-retest reliability of each variables among repetitions at 40, 60, 70, and 80% 1RM for the HPC and HHP. The strength of the correlation was interpreted as slight, fair, moderate, substantial, and almost perfect for ICCs of ≤0.20, 0.21–0.40, 0.41–0.60, 0.61–0.80, and ≥0.81, respectively (10). Two-way repeated-measures analysis of variance (ANOVA) was used to compare peak power, Fpp, Vpp, and peak bar height between the HPC and HHP at each load. When there was a significant interaction, post hoc tests with Bonferroni's correction was used for paired comparisons between the values of the HPC and HHP at each load. Partial eta squared (η2) was reported to illustrate the effect size of the ANOVA tests. Cohen's d effect sizes were calculated and interpreted as follows: 0.00–0.19 (trivial), 0.20–0.59 (small), 0.60 to 1.19 (moderate), 1.20–1.99 (large), 2.00–3.99 (very large), and ≥4.00 (nearly perfect) (5). The significance level was set at p ≤ 0.05. All statistical analyses were performed using SPSS version 25 (IBM Corp, New York, NY).
The ICCs for peak power at each relative load during the HPC and HHP ranged from 0.917 to 0.987, exhibiting almost perfect reliability. The ICCs for Fpp demonstrated almost perfect reliability as well, with ICCs in the range of 0.970–0.995. The ICCs for Vpp were 0.856–0.966 for all loads excluding 60% 1RM for the HHP (0.685, substantial). In addition, the ICCs for peak bar height demonstrated almost perfect reliability, with ICCs ranging from 0.895 to 0.988.
Analysis of variance revealed a significant interaction in peak power between the exercises and loads (p = 0.001, partial η2 = 0.730). There were significant differences between the HPC and HHP in peak power at 40, 60, and 70% 1RM (p = 0.00, 0.003, and 0.002, respectively). Contrarily, no statistical difference was observed between the exercises at 80, 90, 95, and 100% 1RM. The practical significances were large at 40 and 60% 1RM (d = 1.99 and 1.22, respectively) but only trivial to small at 70, 80, 90, 95, and 100% 1RM (d = 0.58, 0.13, 0.04, 0.32, and 0.25, respectively) (Figure 1).
Velocity, Force, and Bar Height
There were significant interactions between the exercises and loads for Vpp (p = 0.001, partial η2 = 0.864), and peak bar height (p = 0.001, partial η2 = 0.911). Although there was a significant exercise main effect on Fpp (p = 0.029, partial η2 = 0.516), no significant interaction was observed. The descriptive values of Fpp, Vpp, and peak bar height at each relative load are listed in Table 1. Peak bar height was significantly greater during the HHP than during the HPC at 40, 60, and 70% 1RM, whereas no significant differences were found at 80, 90, 95, and 100% 1RM (trivial to small practical significance). Similar trends were observed for Vpp, as the differences between exercises decreased at heavier loads and the HPC produced greater values than the HHP at 95 and 100% 1RM.
Table 1 -
The effect of different relative loads on system variables (force and velocity) and bar height during the hang power clean and hang high pull.*
|Load (% 1RM of the HPC)
||2,338 ± 266
||2,655 ± 272
||2,739 ± 296
||2,837 ± 311
||2,878 ± 335
||2,929 ± 381
||2,910 ± 367
||2,539 ± 331
||2,726 ± 289
||2,790 ± 306
||2,879 ± 303
||2,925 ± 353
||2,911 ± 331
||2,969 ± 364
||−201 ± 199
||−71 ± 97
||−51 ± 46
||−42 ± 67
||−47 ± 99
||17 ± 151
||−59 ± 106
||1.35 ± 0.11
||1.43 ± 0.08
||1.42 ± 0.07
||1.41 ± 0.08
||1.35 ± 0.07
||1.33 ± 0.05
||1.31 ± 0.07
||1.61 ± 0.11
||1.58 ± 0.05
||1.48 ± 0.05
||1.40 ± 0.07
||1.32 ± 0.06
||1.28 ± 0.07
||1.24 ± 0.05
||−0.26 ± 0.05
||−0.15 ± 0.07
||−0.06 ± 0.04
||0.01 ± 0.06
||0.03 ± 0.06
||0.05 ± 0.05
||0.07 ± 0.06
|Peak bar height (cm)
||139 ± 5
||132 ± 6
||129 ± 6
||126 ± 7
||121 ± 7
||118.8 ± 6
||117.2 ± 7
||158 ± 8
||146 ± 7
||136 ± 7
||129 ± 6
||122 ± 5
||119.6 ± 5
||116.2 ± 6
||−21 ± 9
||−14 ± 5
||−7 ± 6
||−3 ± 6
||−1 ± 4
||−1 ± 4
||1 ± 4
*HPC = hang power clean; HHP = hang high pull; 1RM = 1 repetition maximum; Fpp = force at peak power; Vpp = velocity at peak power; diff = difference between the exercises calculated as the value of the HPC minus that of the HHP; d = Cohen's d effect size between the HPC and HHP; n.s. = not significant.
†Significant difference between the HPC and HHP.
This study aimed to compare power output between the HPC and HHP across the wide range of loads among weightlifters. The peak power produced during the second pull phase was significantly greater for the HHP than for the HPC at 40, 60, and 70% 1RM (small to large practical significance). However, there was no statistical difference in peak power between the exercises at 80, 90, 95, and 100% 1RM (trivial to small practical significance), which supports our hypothesis (Figure 1).
At 40, 60, and 70% 1RM, for which significant differences in peak power were observed, the peak bar height was significantly greater for the HHP than for the HPC (d = 1.01–3.11) (Table 1). This suggests the necessity of voluntarily limiting the bar height in the HPC at lighter loads because of the catching movements, which likely hindered the subjects from lifting with their maximal efforts. This notion is supported by the fact that during the HPC, bar height did not differ at lighter loads, whereas it significantly decreased at heavier loads (21). Because there is a strong positive correlation between velocity and height in the HPC (21), the subjects did not produce maximal force and velocity in the HPC at 40–60% 1RM, resulting in the significantly lower peak power compared with that of the HHP (Figure 1). Contrarily, at heavier loads (80–100% 1RM), there was no significant difference in peak bar height between the exercises (d = 0.12–0.47) (Table 1). That is, subjects could execute the second pull movements with their maximal efforts at 80% 1RM or greater, which allowed them to generate equivalent peak power in the HPC and HHP. The HPC produced a greater Vpp than the HHP at 95 and 100% 1RM. The differences in velocity at maximal loads possibly arose from the psychological pressure of catching the bar during the HPC, although the instruction for maximal efforts was given in both the HPC and HHP sessions. These results suggest that the previous study reported the greater power output in the HHP than in the HPC across the entire loading spectrum partly because the investigated loading range (30–80% of 1RM) was not wide enough (20). The current study used wider range of loads including loads >80% 1RM, and showed that the differences in peak power between the exercises diminished toward heavier loads.
In addition, in previous studies (18–20), it was suggested that catching derivatives produced less power than pulling derivatives because the intention to receive the bar may have caused an incomplete triple extension during the second pull phase. Based on the results of the current study, the incomplete triple extension may have arisen from the lifting technique of the subjects opposed to the characteristics of the catching derivatives. Unlike the previous studies using nonweightlifters (e.g., competitors in NCAA Division III track and field or collegiate club/intramural sports) (18–20), the current study used competitive weightlifters to compare pure power characteristics between the exercises. They could most likely perform the HPC with the correct technique, even with heavier loads, because near-maximal to maximal loads are often used in their daily training and competitions. The movements up to the second pull phase, in which peak power occurs, are similar between the HPC and HHP. Thus, it is suggested that both exercises produce extremely high and comparable power regardless of the presence/absence of the catching phase, if done correctly. For power development, the HHP might be a superior exercise over the HPC when using 40–70% 1RM, whereas the exercises can be interchangeably used with loads of 80–100% 1RM.
To the best of our knowledge, there is only one published intervention study of the effect of training using catching or pulling derivatives (1). In the study, subjects were divided into 2 groups, with one using the power clean and the other using the clean pull. After 8 weeks of training, both groups displayed improvements in most performance variables for jumping and isometric midthigh pull, but there was no significant difference in the percentage increase after the interventions between the groups. This was contradictory to their initial hypothesis that the pull group would experience a greater improvement than the catch group. From the results of the current study, the possible explanation for the lack of differences between the groups is the load used for the training intervention. The current study used a wider range of loads (40–100% 1RM) than previous studies (30–80% 1RM) (18–20) and observed no significant difference in peak power between the HPC and HHP at 80% 1RM or greater. The aforementioned intervention study used loads of ≥80% 1RM during half of the sessions, which may have led to similar training effects on the performance improvements. Although it is unclear whether the subjects performed those movements correctly during the training sessions, the difference in peak power between the HPC and HHP tended to decrease at heavier loads in both the current study (Figure 1) and previous studies (19,20). Thus, it seems that there is no clear difference in training effects between the exercises if similar and heavier loads are used. What should be remembered is that pulling derivatives allow lifters to use loads heavier than 1RM of catching derivatives, making it possible to provide training stimuli on force-velocity profiles differently than catching derivatives (1,15). Therefore, further research is needed to verify the mechanism by which using different loads for the catching and pulling derivatives results in power developments and improvements among athletic performances.
A limitation of the current study was the inclusion of weightlifters as the subjects. The results of this study might not be directly applicable to other populations who perform weightlifting and its derivatives as power training. However, it was valid to adopt them as subjects because the aim of this study was to compare the pure power characteristics of the movements while eliminating the effect of lifting technique.
It has been recommended that to maximize power output during athletic performances, athletes should train at a wide range of loads with exercises that produce greater power (4,7,11). In a previous study, the HHP produced greater power than the HPC for all loads investigated, suggesting that the HHP provides a superior training stimulus (20). However, the results of the current study suggest that the differences in power output between the exercises diminish toward heavier loads if done correctly. Thus, when training with weightlifting derivatives, coaches and athletes should select appropriate exercises depending on the intended loads: at 40–70% 1RM the HHP should be selected over the HPC, whereas at 80% 1RM or greater, the HPC and HHP can be interchangeably used as effective exercises for power development.
The authors thank all subjects for their cooperation.
This work was supported by JSPS KAKENHI Grant Number JP17K01696 and a Waseda University Grant for Special Research Projects (project number: 2019C-367). The results of this study do not constitute an endorsement of the product by the authors or the journal. None of the authors declared a conflict of interest.
1. Comfort P, Dos'Santos T, Thomas C, McMahon JJ, Suchomel TJ. An investigation into the effects of excluding the catch phase of the power clean on force-time characteristics during isometric and dynamic tasks: An intervention study. J Strength Cond Res 32: 2116–2129, 2018.
2. Cormie P, McCaulley GO, Triplett NT, McBride JM. Optimal loading for maximal power output during lower-body resistance exercises. Med Sci Sports Exerc 39: 340–349, 2007.
3. Garhammer J. A review of power output studies of Olympic and powerlifting: Methodology, performance prediction, and evaluation tests. J Strength Cond Res 7: 76–89, 1993.
4. Haff GG, Nimphius S. Training principles for power. Strength Cond J 34: 2–12, 2012.
5. Hopkins WG. A Scale of Magnitude for Effect Statistics. 2006. Available at: https://www.sportsci.org/resource/stats/effectmag.html
. Accessed September 3, 2019.
6. Hori N, Newton R, Nosaka K, Stone M. Weightlifting exercises enhance athletic performance that requires high-load speed strength. Strength Cond J 27: 50–55, 2005.
7. Kawamori N, Haff GG. The optimal training load for the development of muscular power. J Strength Cond Res 18: 675–684, 2004.
8. Kipp K, Comfort P, Suchomel TJ. Comparing biomechanical time series data during the hang-power clean and jump shrug. J Strength Cond Res 2019. Epub ahead of print.
9. Kipp K, Malloy PJ, Smith JC, et al. Mechanical demands of the hang power clean and jump shrug: A joint-level perspective. J Strength Cond Res 32: 466–474, 2018.
10. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 33: 159–174, 1977.
11. Newton RU, Kraemer WJ. Developing explosive muscular power: Implications for a mixed methods training strategy. Strength Cond Res 16: 20–31, 1994.
12. Oranchuk DJ, Drinkwater EJ, Lindsay RS, et al. Improvement of kinetic, kinematic, and qualitative performance variables of the power clean with the hook grip. Int J Sport Physiol Perform 14: 378–384, 2019.
13. Stone MH, Pierce KC, Sands WA, Stone ME. Weightlifting: Program design. Strength Cond J 28: 10, 2006.
14. Storey A, Smith HK. Unique aspects of competitive weightlifting: Performance, training and physiology. Sports Med 42: 769–790, 2012.
15. Suchomel TJ, Comfort P, Lake JP. Enhancing the force-velocity profile of athletes using weight-lifting derivatives. Strength Cond J 39: 10–20, 2017.
16. Suchomel TJ, Comfort P, Stone MH. Weightlifting pulling derivatives: Rationale for implementation and application. Sports Med 45: 823–839, 2015.
17. Suchomel TJ, Lake JP, Comfort P. Load absorption force-time characteristics following the second pull of weightlifting derivatives. J Strength Cond Res 31: 1644–1652, 2017.
18. Suchomel TJ, Sole CJ. Force-time-curve comparison between weight-lifting derivatives. Int J Sport Physiol Perform 12: 431–439, 2017.
19. Suchomel TJ, Sole CJ. Power-time curve comparison between weightlifting derivatives. J Sports Sci Med 16: 407–413, 2017.
20. Suchomel TJ, Wright GA, Kernozek TW, Kline DE. Kinetic comparison of the power development between power clean variations. J Strength Cond Res 28: 350–360, 2014.
21. Takei S, Hirayama K, Okada J. Is the optimal load for maximal power output during hang power cleans sub-maximal? Int J Sports Physiol Perform 15: 18–24, 2020.