Explosive upper extremity movements are used during sports such as basketball and baseball, occupational activities such as postal service employees pushing boxes around or a housekeeper at a hotel throwing laundry, and for rehabilitation purposes (2,5,10–12,14,17). Much research today includes plyometric movements involved with lower body activities such as box drop jumps (1,4,9,14,21,22), ankle hops (9,21,22), jump squats (7,14,21,22), vertical (13), and horizontal jumps (4,9,14,21–23), single leg hops (3,9,21,22), bounding (21,22), tuck jumps (14,21,22), and stop jumps (23). Although limited research studies have examined training adaptations associated with upper extremity plyometrics (2,12,17,20), even fewer have quantified the loads and characteristics associated with many of the specific exercises.
The majority of research on upper body plyometrics has focused on open kinetic chain power production and power development adaptations associated with medicine ball chest passes (9), bench press ballistic exercises (11,16), medicine ball power drops (7), and overhead throwing movements (2,8,11,12,14,17,20). One study considered spinal loading during closed kinetic chain push-ups, such as stationary push-ups and staggered push-ups, to different forms of plyometric push-ups (medicine ball and clap push-ups) (10). Thus, there is a paucity of research examining upper extremity explosive plyometric push-ups, particularly with respect to the magnitude and temporal characteristics imposed upon the upper extremity.
When designing a plyometric program, intensity may be the most important factor to consider (14). Intensity levels can be determined through the examination of the magnitude and temporal characteristics of vertical ground reaction forces (vGRF) such as peak forces, time to peak forces, propulsion rates (PRs), loading rates (LRs), and ground contact time (GCT) (3,9,14). When comparing lower extremity exercise intensities by vGRF, it can be assumed that a low peak vGRF, <2 times the individual's body weight, reflects a low-intensity exercise, whereas a high vertical GRF, ≥4 times the individual's bodyweight, represents a more advanced exercise (19,22). It is important to evaluate the stresses imposed with different plyometric exercises so that a progressive program can be designed for the participant.
The purpose of this study was to compare the vGRF characteristics between 4 different plyometric push-up variations and limbs (dominant, nondominant) to accurately define intensity. The 4 push-up variations included the clap push-up, box drop push-up from 3.8 cm (BD1), BDP from 7.6 cm (BD2), and BDP from 11.4 cm (BD3). We hypothesized that there would be an intensity hierarchy from lowest to highest, BD1, BD3, clap, BD4, on the produced levels of GRF.
Experimental Approach to the Problem
Vertical ground reaction forces during 4 upper extremity plyometric push-up exercises were measured in a sample of 22 men to determine the exercise intensity. All data were collected during a single session, with push-up variation order randomized between subjects. Specific GRF characteristics included dominant and nondominant limb GCT, peak force, time to peak force, LR, and PR. Independent variables were push-up variation (clap push-up, BD1, BD2, and BD3) and limb (dominant, nondominant).
Twenty-two physically active men were recruited for this study (mean ± SD age 25.9 ± 1.3 years, weight, 87.6 ± 12 kg, and height, 1.80 ± 0.08 m). All the participants had no history of a diagnosed upper extremity injury (chronic or acute) or surgery within the past 6 months. Physically active was defined as working out for a minimum of 3 d·wk−1, 20-minute sessions, for 3 consecutive months. All the participants included upper body strength training at least 1 d·wk−1, including exercises similar to the bench press, push-ups, or medicine ball push-ups. This study had institutional review board approval with each participant signing an approved institutional review board informed consent document before study participation.
A calisthenic systemic warm-up preceded the exercises that consisted of 10 repetitions of bodyweight squats, static lunges, and jump squats. Once the warm-up was completed, the participants were instructed to perform the 4 plyometric push-up variations. Before any data collection, the subjects were allotted time to practice the exercises until they felt comfortable with each variation and qualitative observation yielded proper technique. This also served as an upper extremity warm-up before actually performing the test repetitions. The order in which the variations were performed when data were collected was randomized for each subject.
For all 4 variations, the subjects placed each hand on a separate force plate using a self-selected hand placement width. The starting position for each push-up variation consisted of the participant in the “up” phase of a push-up with arms fully extended, torso in a straight line from their head to their heels, heels up with body weight on toes, and neck extended (Figure 1).
During the clap push-ups, the participants lowered their chest, down toward the force plates, while maintaining a straight line with their body. After this phase, they would forcefully push up and while in the air clap their hands together before landing back onto the force plates. One full repetition was defined as successfully completing both phases of the exercise and returning to the starting position after landing back onto the force plates.
The 3 box drop plyometric push-ups were performed using 3.8-cm wooden blocks. The BD1 used 1 block (3.8 cm), BD2 2 blocks stacked on top of each other (7.6 cm), and BD3 used 3 blocks stacked (11.4 cm). The participants began with their hands on the force plates, with the blocks placed just outside their self-selected hand placement width. From this position, similar to the clap push-ups, they lowered their chest down toward the force plates while maintaining their body in a straight line. Once peak depth was obtained, they explosively pushed up and while in the air maneuvered their hands such that they landed on the blocks simultaneously. Upon contact with the blocks, the subjects immediately explosively performed another push-up off the blocks, landing with their hands in the starting position (on the force plates) to perform a subsequent repetition. Thus, the BDPs consisted of 2 phases. The first phase (Figure 2) began when the hands made contact with the force plate after the push-off from the blocks. The first phase concluded when the propulsion effort to propel the body into the air so that the hands could land back onto the boxes ended (hands no longer in contact with the force plates). The second phase involved the part of the push-up when the hands were in contact with the blocks (Figure 3). Only the GRF data collected during the first phase, impact and propulsion from the force plates, were analyzed.
Although the pace of each exercise was self-selected, the subjects were instructed to perform each repetition consecutively with maximal explosiveness and no pause (rest) between repetitions. After the completion of the 4 repetitions of each set, the subjects had a self-determined break of not <90 seconds before they began the next randomized variation set of push-ups.
Instrumentation and Data Reduction
Ground reaction forces were collected using 2 force plates (BP400600NC 2000 Advanced Mechanical Technology, Inc., Watertown, MA, USA) such that each force plate recorded the ground reaction forces separately for each extremity. Data from the force plates were collected (1,000 Hz) and converted to GRF data using the Motion Monitor acquisition software package (Innovative Sports Training, Inc: Chicago IL, USA). Data reduction was conducted offline using Matlab (The Mathworks, Inc., Natick, MA, USA) based scripts to compute 5 force and temporal variables. First, GRFs were filtered using a zero phase lag Butterworth filter (50-Hz cutoff). Ground contact time was defined as the length of time that the total vGRF (sum of vGRF under the dominant and nondominant extremities) was >15 N. The remaining dependent variables were computed separately using the GRFs recorded under the dominant and nondominant extremities. Peak vGRFs recorded during the time of ground contact were computed and normalized to body weight (kg × 9.8 m·s−1). The time to reach peak vGRF was computed as the time between ground contact and when the peak vGRF was attained. The LR was computed as the slope between when the vGRF >50–50 N plus one-third body weight. The PR was computed as the slope between when the vGRF decreased <50 N plus 1 quarter body weight to <50 N. The thresholds for the end and beginning of LR and PRs, respectively, were set below thresholds normally used for lower extremity impacts because full, unsupported body weight would not be sustained in a push-up position. The exact values used, one-third and one-quarter body weight, were established following a review of the first 5 subject ensemble averages.
All statistical analyses were performed using SPSS for Windows (SPSS, Inc. Chicago, IL, USA). Intraclass correlation coefficients and SEM were computed to determine the reliability of each dependent variable across the multiple repetitions. using a 1-way repeated analysis of variance (ANOVA) for ground contract time and separate 2 way (variation by limb) repeated measures ANOVA for peak vertical GRF, time to peak vertical GRF, LR, and PR. Exploratory analyses were conducted to examine the data with regard to normality. When issues of sphericity were revealed, degrees of freedom were adjusted using Huynh-Feldt corrections. Bonferroni-adjusted pairwise post hoc comparisons were made when appropriate. Effect sizes for pairwise comparisons were calculated as Hedges's g (15). Statistical significance was set at an α level of 0.05.
Results of the intraclass correlation coefficients and SEM demonstrated a moderate to high reliability (Table 1). Descriptive statistics and 95% confidence intervals for GCT are presented in Table 2, whereas the remainder of the dependent variables is presented in Table 3. Ground contact time for the clap push-up (F 1.9, 40.6 = 3.75; p = 0.033) was significantly less than BD2 (p = 0.033, d = 0.43) and BD3 (p = 0.001, d = 0.59).
Across all 4 variations, the dominant limb peak vertical GRF was significantly greater than the nondominant limb (F 1,21 = 4.52; p = 0.045, d = 0.16). There were no significant peak vGRF differences between conditions based on the condition main effect (F 3,63 = 2.72; p = 0.52) and variation by limb interaction (F 3,63 = 1.32; p = 0.0.276).
There was no significant difference for time to peak force between limbs (F 1,21 = 0.14; p = 0.717), variation (F 3,63 = 2.43; p = 0.074), or limb by variation (F 2.3,48.6 = 1.32; p = 0.278).
There was a significant variation by limb interaction for LR (F 3,63 = 25.55; p < 0.001). Within limb between push-up variation comparisons for LR are summarized in Table 4. Additionally, post hoc comparisons yielded the dominant limb LR to be significantly greater than that for the nondominant for the clap push-up (p < 0.001, d = 1.83), BD1 (p < 0.001, d = 0.65), BD2 (p < 0.001, d = 0.15), and BD3 (p < 0.001, d = 0.87).
For PR (F 3,63 = 11.56; p < 0.001), the clap push-up was significantly greater than BD1 (p < 0.001, d = −0.79), BD2 (p < 0.001, d = −0.70), and BD3 (p < 0.001, d = −0.84). There were no differences for the PR between limbs based on the limb main effect (F 1,21 = 0.37; p = 0.550) or variation by limb interaction (F 3,63 = 0.69; p = 0.561).
The purpose of this study was to compare selected vGRF characteristics between 4 different plyometric push-up variations, the clap push-up, and BDPs from 3.8, 7.6, and 11.4 cm. The most remarkable result was the absence of significant peak vGRF magnitude or timing differences between exercises. In addition, contrary to our hypothesis, the clap push-up appears to have the highest loading and PRs along with shortest GCT. Based on the loading and PR results, it appears that the clap push-up is the exercise with the greatest intensity followed by box drops from 11.4 cm.
The LR reflects the early stages of eccentric loading immediately after ground contact before active neuromuscular modulation of the impact shock could occur (18). Ricard and Veatch describe LR as occurring at 50–75 milliseconds into initial impact and untrainable in individuals. However, time to peak vertical GRF during all 4 push-up variations occurred much later than the LR explaining why peak vGRF may not have shown a significant difference between variations. Time to peak vGRF ranged between 410 and 460 milliseconds after impact, giving the body more time to adjust joint kinetics and kinematics for a softer landing. Santos-Rocha et al. (19). reported the LR to be directly related to step rate pace, whereas GCT was inversely related to step rate pace. This likely explains the combination of the significantly shorter GCT and significantly higher LR for the clap push-up compared with the BDPs.
Surprisingly, there were no significant differences in peak vGRF between the push-up variations. The rationale for our hypothesis was the assumption that there would be peak flight height differences between the variations, with ground contact velocity being directly related to peak flight height. Greater ground contact velocity is associated with greater momentum at ground contact, which would subsequently require changes in the force magnitude and application time to reverse the downward body displacement in preparation for the next repetition. In addition to our assumption of peak flight differences being potentially incorrect, the, lack of significant peak vGRF differences may also be attributed to the subjects modifying their landing strategies. Specifically, the subjects may have subconsciously adjusted their shoulder, elbow, and wrist flexion when landing, thereby creating a softer landing, which corresponded to peak vGRF differences between the push-up variations. Softer landings have been shown to produce lower vGRF values during lower extremity movements such as the drop landings. According to DeVita and Skelly (6), the kinematics of the ankle, knee, and hip were different between stiff and soft drop landings. As flexion increases in the joints during a landing, vGRFs are absorbed over a greater length of time by eccentric muscle action, thus decreasing the peak impact forces. Although DeVita and Skelly refer to lower extremity impacts and their effects on musculoskeletal tissues, there is no reason that the same concept for lower body plyometrics does not apply for upper extremity plyometrics (17,21).
Additionally, the peak forces measured during the 4 push-up variations demonstrated that the peak loading the musculoskeletal system experienced during impact under each hand was approximately 0.7 times body weight, for a total of 1.4 times the body weight across both extremities. Again, the lack of statistical significance between variations was an unexpected result because of the assumed peak flight height differences of each push-up variation. It is not surprising that the plyometric push-up forces studied in the current investigation are less than the forces associated with lower extremity plyometrics because lower extremity plyometrics involve the entire body whereas plyometric push-ups involve approximately three-quarters of the body. This creates less impact forces during the landing because of less weight being propelled into the air and being involved with the ground contact. Additionally, the height of drop for the variations we studied are less than the height of drop associated for many lower extremity plyometrics, which might also account for the peak vGRF being less than lower extremity plyometrics. According to Witzke and Snow (22), low-intensity plyometrics for the lower extremity can be defined as a peak vGRF 2 times the body weight or less and that a minimum peak vGRF of 1–2 times the body weight corresponds to osteogenic effects. Based on the upper extremity not supporting body weight as regularly as the lower extremity, we can assume the intensity levels based on vGRF for lower body plyometrics does not apply to the upper extremity plyometrics in regard to what is needed to stimulate adaptations such as osteogenesis. Thus, although peak vGRF 2 times the body weight is considered to be low intensity for the lower extremity, it might be considered to be mid to high intensity for the upper extremity. Once further research has further investigated the vGRF associated with other types of plyometric exercises, recommendations can be made regarding a new intensity scale used solely for describing upper extremity plyometrics.
Even though no significant bilateral differences revealed for PR between push-up variations, there was bilaterally asymmetry with the dominant limb demonstrating significantly greater peak forces than the nondominant limb. A disparity between limbs may be attributable to the dominant limb being stronger and having more neural adaptations in regards to handling higher loads. This might explain the dominant and nondominant differences during LR as well. Future research should consider studying the relationship between upper extremity strength and the impact forces during plyometric push-ups.
Propulsion rate demonstrated the biggest differences between the 4 push-up variations. Propulsion rate (concentric force production phase) is similar to the LR (eccentric deceleration phase) because it represents a change in the force being produced during a specific time. The difference between the 2 rates is that PR (concentric) occurs during the terminal ascent phase, just before ground off, whereas LR occurs during the early descent phase (eccentric phase), when the body makes contact with the ground. Furthermore, in addition to LR largely being influenced by ground contact momentum, LR is often associated with injury potential because it involves eccentric muscle action during a period of time after ground contact before active modulation by the neuromusculoskeletal system can occur (19). In contrast, PR represents the terminal results of voluntary, concentric force production. Thus, because it is a function of voluntary muscle activation, it is more representative of muscle power without the higher injury risk associated with the loading phase.
As with many studies, a limitation of this study included the accessible population from which subjects were recruited. One unique aspect of our study was activity level status of the subjects. Slightly over half of our subjects were active duty Marines who performed a variety of push-ups on a regular basis, whereas the remainder of the subjects met the stated inclusion criteria but may not have participated in push-up exercises as regularly as the Marines did. Furthermore, the push-ups were completed on 2 fiberglass force plates vs. a more compliant surface such as aerobics floor. The fiberglass force plates may have led the subjects to believe that the impact would hurt, whereas if they had performed the push-up variations on a more compliant floor, they may have landed differently. In addition, no constraints on hand placement were used during the push-ups. This may have increased the possibility of hand-width variations, thereby causing changes in the kinetics and kinematics of the upper extremity, which in turn could have increased the variability of the peak GRF and temporal characteristics.
Further research is encouraged to fill the gap that exists between lower extremity and upper extremity plyometric knowledge. Future studies may wish to evaluate kinematic and kinetic data to determine if joint angles are an integral part in the GRF data. It may also be interesting to standardize hand and elbow placement and see if this affects any of the independent variables that are currently discussed. Moreover, developing an intensity scale for upper extremity plyometrics would be valuable for practicioners to apply the principles of progression and overload in the design of an upper extremity plyometric training program.
In using plyometrics for strength and conditioning, intensity may be the most important factor, but to date, little is known about the intensity of upper extremity plyometric exercises. The major practical applications of these results are the quantification of loading the upper extremity experiences during 4 upper extremity plyometric push-up variations. First, it is important for the strength coach to understand that the peak loading experienced by the musculoskeletal system of each upper extremity was 0.7 times the body weight (total of 1.4 times body weight across both extremities). This can be used in computing the total loading an athlete is subjected to during a workout to ensure sufficient stimulus for adaptation while avoiding overtraining. This study demonstrated symmetrical loading across all 4 plyometric push-ups, with the dominant limb demonstrating significantly greater peak impact forces. In addition, LRs for each of the push-up variations also demonstrated greater values for the dominant limb. Although future research might reveal the etiology of underlying loading asymmetry to assist with program design, strength coaches should be aware that asymmetrical loading appears to occur during plyometric push-ups. Based on the significantly higher PRs and shorter GCTs, it appears that the clap push-up is the most intense exercise. Although not all values were statistically significant, changing the BDP heights by 3.8 cm appears to be an adequate stimulus for progressively increasing exercise demands. The lack of potent peak force or time to peak force differences between the push-up variations studies suggests that further research is needed to examine upper extremity kinematics of plyometric push-ups.
1. Blackburn J, Pagua D. Sagittal plane trunk position, landing force, and quadriceps electromyographic activity. J Athl Train 44: 174–179, 2009.
2. Carter A, Kaminski T, Douex A, Knight C, Richards J. Effects of high volume upper extremity
plyometric training on throwing velocity and functional strength ratios of the shoulder rotators in collegiate baseball players. J Strength Cond Res 21: 208–215, 2007.
3. Coventry E, O'Connor K, Hart B, Earl J, Ebersole K. The effect of lower extremity fatigue on shock attentuation during single leg landing. Clin Biomech 21: 1090–1097, 2006.
4. Crowther R, Spinks W. Kinematic responses to plyometric exercises conducted on compliant and non-compliant surfaces. J Strength Cond Res 21: 460–465, 2007.
5. Davies G, Matheson D. Shoulder plyometrics. Sports Med Arthrosc 9: 1–18, 2001.
6. Devita P, Skelly W. Effect of landing stiffness on joint kinetics and energetics in the lower extremity. Med Sci Sports Exer 24: 108–115, 1992.
7. Ebben W, Blackard D, Jensen R. Quantification of medicine ball vertical impact forces: Estimating effective training loads. J Strength Cond Res 13: 271–274, 1999.
8. Ebben W, Blackard D, Jensen R. Electromyographic and kinetic analysis of complex training variables. J Strength Cond Res 14: 451–456, 2000.
9. Ebben W, Simenz C, Jensen R. Evaluation of plyometric intensity using electromyography. J Strength Cond Res 22: 861–868, 2008.
10. Freeman S, Karpowicz A, Gray J, McGill S. Quantifying muscle patterns and spine load during various forms of the pushup. Med Sci Sports Exer 38: 570–577, 2005.
11. Frost D, Cronin B, Newton R. A novel approach to identify the end of the concentric phase during ballistic upper-body movements. J Strength Cond Res 24: 282–286, 2010.
12. Heiderscheit B, McLean K, Davies G. The effects of isokinetic vs plyometric training on the shoulder internal rotators. J Orthop Sports Phys Ther 23: 125–133, 1996.
13. Jensen R, Ebben W. Kinetic analysis of complex training rest interval effect on vertical jump performance. J Strength Cond Res 17: 345–349, 2003.
14. Jensen R, Ebben W. Quantifying plyometric intensity via rate of force development, knee joint, and ground reaction forces. J Strength Cond Res 21: 767, 2007.
15. Kline R. Beyond Significance Testing. Washington DC: American Psychological Association, 2004.
16. Newton R, Murphy A, Humphries B, Wilson G, Kramer W, Hakkine K. Influence of load and stretch shortening cycle
on kinematics, kinetics, and muscle activation that occurs during explosive
upper-body movements. Eur J Appl Phys 75: 333–342, 1997.
17. Nicol C, Avela J, Komi P. The stretch shortening cycle
: A model to study naturally occurring neuromuscular fatigue. Sports Med 36: 977–999, 2006.
18. Ricard M, Veatch S. Comparison of impact forces in high and low impact aerobic dance movements. J Appl Biomech 6:67–77, 2009.
19. Santos-Rocha R, Veloso A, Machado M. Analysis of ground reaction forces in step exercise depending on step pattern and stepping rate. J Strength Cond Res 23: 209–224, 2009.
20. Schulte-Edelmann J, Davies G, Kernozek T, Gerberding E. The effects of plyometric training of the posterior shoulder and elbow. J Strength Cond Res 19: 129–134, 2005.
21. Vescovi J, Canavan P, Hasson S. Effects of a plyometric program on vertical landing force and jumping performance in college women. Phys Ther Sport 9: 185–192, 2008.
22. Witzke K, Snow C. Effects of plyometric jump training on bone mass in adolescent girls. Med Sci Sports Exer 32: 1051–1057, 2000.
23. Yu B, Lin C, Garrett W. Lower extremity biomechanics during the landing of a stop-jump task. Clin Biomech 21: 297–305, 2006.
Keywords:© 2012 National Strength and Conditioning Association
stretch shortening cycle; upper extremity; explosive; push; chest