Maximal power output is produced when an optimal level of both force and velocity is achieved (28). For example, during a ballistic exercise, if the external load is too large, velocity may be compromised; whereas, if the external load is too light, force may be compromised, and in both instances, result is suboptimal power production. The load that maximizes power has been repeatedly assessed (2,5–7,9–11,16,17,20,25), with large variation in the optimal load reported both between separate investigations and also between subjects within investigations. Specifically, for the upper body, a range of 15–70% one repetition maximum (1RM) has been suggested to produce maximal power (5,6,9,12,25). The variation in findings is probably because of differences in methodologies used (9) and subject characteristics (6,28).
Bevan et al. (9) highlighted that the large variation in the optimal load was probably because of methodological differences, such as data collection techniques, exercises performed, strength levels of the subjects, and the reporting of either mean or peak power. To standardize the methodological design, Bevan et al. (9) determined that a protocol which included a ballistic bench press, where peak power was recorded via a linear position transducer in strength trained athletes (professional Rugby Union players), would allow better comparison with previous literature. It was reported that a load equaling 30% of 1RM allowed maximal power to be produced (9). However, by assessing the load that produces maximal power using a different method to that previously outlined in the literature, Bevan et al. (9) may have added to the variation in findings reported. Indeed, the study by Baker et al. (6) recruited trained athletes, but assessed mean power output, and reported that maximal power was produced at 55% of 1RM. Whereas, Newton et al. (25) assessed peak power, but recruited sport science students, and reported that maximal power was produced at 15 and 30% of 1RM. As such, to minimize variation in the load that produces maximal power reported between studies, methodologies need to be replicated.
Regardless of methods used, authors have also reported a considerable between-subject variation in the load that maximizes power within studies that may be due to differences in subject characteristics. For example, Baker et al. (6) reported SD of 5.3% on either side of the mean bench throw power output (54.9% of 1RM) in rugby league players. If one were to extrapolate findings to include the whole population tested (i.e., 3 SDs or 99.74% of subjects), maximal power production would range between 39 and 71% 1RM. Similarly, Bevan et al. (9) reported that while 30% 1RM produced maximal power for the entire group, there were no significant differences between power outputs at 30, 40, or 50% 1RM. Relationships between the load that produces maximal power and maximal strength, limb length, or girth measurement may help to identify some of the between-subject variation. Indeed, Baker (5) reported that stronger athletes use a lower percentage of 1RM to produce maximal power compared with their less-strong counterparts. It has also been reported that athletes with more muscular limbs may have greater lifting capacity (18). However, the effect of anthropometric characteristics on the load that produces maximal power has not been established.
Therefore, the aim of this investigation was to determine the extent in which anthropometric measures correlate with the load that produces maximal power. To minimize variation in findings between studies, the methods outlined by Bevan et al. (9) were used.
Experimental Approach to the Problem
To determine the extent in which anthropometric measures correlate with the load that produces maximal power, professional Rugby Union players were assessed for upper-body maximal strength and power across 3 sessions. During the first session, players were assessed for bench press strength and anthropometric measures (body mass, height, upper-arm girth, upper-arm length, and forearm length). During the second and third sessions, players were assessed for peak power during a ballistic bench throw at a range of loads (20, 30, 40, 50, and 60% 1RM; Table 1). Power testing was split into 2 separate sessions to limit fatigue. Each session was separated by 24 hours. Peak power measured in a bench throw with a linear position transducer was selected in an attempt to replicate the assessment methods in the study by Bevan et al. (9).
Twenty-six elite rugby union players from a Super 14 professional Rugby Union team during the preseason phase of their campaign volunteered to take part in this study (mean ± SD: age, 24.0 ± 2.3 years; height, 183.6 ± 2.9 cm; mass, 101.1 ± 10.1 kg). Each player had undergone at least 2 years of intensive and regular resistance training exercise and was well familiarized with all the testing procedures. Players were informed of the experimental risks and signed an informed consent document before the investigation. The investigation was approved by an Institutional Review Board for use of human subjects (Auckland University of Technology Ethics Committee).
International Society for the Advancement of Kinanthropometry protocols were used to determine the anthropometric profile of the rugby players (27). Measurement included body mass, standing height, upper-arm girth, upper-arm length (acromiale-radiale length), and forearm length (radiale-stylion length). Total arm length was considered to be the sum of upper-arm and forearm length.
Maximal strength was assessed using the bench press exercises using methods previously outlined (1,3). Briefly, after 3 submaximal sets of bench press (approximately 60, 80, and 90% 1RM), each player performed 1 set to failure of 1 to 4 repetitions. The bench press was performed using free weights. Three-minute rest was allowed between each set. The final set was used to predict the players' 1RM.
The following equation was used to predict bench press 1RM (21). This equation is a valid measure of 1RM strength as it has been shown to have a correlation between actual and predicted 1RM of r = 0.993 (22):
Upper-body power was assessed using a bench throw exercise performed in a Smith machine. Players warmed up with 1 set of 10 repetitions of press-ups, 1 set of 5 repetitions of clap press-ups, and 2 sets of 3 repetitions of bench throw at 30 and 50% of their 1RM bench press. Players then completed 1 set of 4 repetitions of the bench throw at 20, 30, 40, 50, or 60% 1RM bench press. Players used a self-selected hand position and were instructed to lower the bar to approximately a 90° angle at the elbow. Players were then required to throw the bar vertically and explosively as possible trying to propel the bar for maximum height (3,24). After attainment of maximal height, and on the downward phase, the bar was caught by 2 “spotters” and lowered gently into the subject's hands. Five minutes of rest separated each set.
The power produced during each bench throw repetition was quantified with a Gymaware optical encoder (50 Hz sample period with no data smoothing or filtering; Kinetic Performance Technology, Canberra, Australia) using the methods described elsewhere (13). Quantification of the power produced during the bench throw was calculated using the bar mass (14).
To estimate the load that maximized mechanical power output in the upper body, a quadratic was fitted to each player's power output (in Watts) and load (% of 1RM). The goodness of fit of the quadratic was expressed as an overall correlation coefficient (R 2) calculated by subtracting the total variance from the error variance to determine the variance explained by the model. The variance explained was then divided by the total variance to give R 2. Findings were discussed as means and SDs.
In addition to fitting a quadratic, standardized differences of the mean were used to assess magnitudes of effects between each individual load assessed by dividing the differences by the appropriate between-athlete SD. Standardized changes of <0.19, 0.19–0.59, 0.60–1.19, 1.20–1.99, and >2.0 were interpreted as trivial, small, moderate, large, and very large effects, respectively (8,19). To make inferences about the true (large sample) value of an effect, the uncertainty in the effects were expressed as 90% confidence limits. An effect was deemed unclear if its confidence limits overlapped the thresholds for both the smallest positive and the negative effects.
Pearson's correlation coefficients (and 90% confidence limits) were also calculated between a subject's % 1RM that produced maximal power, and strength and anthropometric measures. Correlations of <0.09, 0.10–0.29, 0.30–0.49, 0.50–0.69, 0.70–0.89, 0.90–0.99, and 1.00 were interpreted as trivial, small, moderate, large, very large, nearly perfect, and perfect, respectively (19). An effect was deemed unclear if its confidence limits overlapped the thresholds for both the smallest positive and the negative correlations.
The interclass correlation and coefficient of variation for bench throw at 30 and 50% of 1RM bench press were assessed in our laboratory using a similar population and were 0.932 and 5.9% and 0.900 and 5.0%, respectively. All test-retest reliabilities were assessed 7 days apart. Validity of the Gymaware optical encoder has been previously reported elsewhere (13). The sample size for this investigation was limited to the number of players in the squad. All players in the squad were included and therefore no more players could be obtained.
The mean predicted 1RM bench press was 121.7 kg (±19.2 kg). The estimated upper-body peak power was 1298 W (±292 W) and occurred at 30.2% (±13.6%) 1RM bench press (mean goodness of fit R 2 = 0.72) (Figure 1). Peak power at 30% of 1RM bench press was found to result in greater (small-to-moderate effect size) peak power than all other intensities assessed. Of the 5 loads assessed, 19 of 26 players produced the greatest upper-body peak power at 30% 1RM bench press.
Correlations between the percentage of 1RM that produced peak power and upper-arm length, forearm length, total arm length, and upper-arm girth were r = −0.61 (90% confidence limits −0.35 to −0.78), −0.29 (0.04 to −0.57), −0.56 (−0.28 to −0.75), and −0.29 (0.04 to −0.57), respectively. The largest relationship of r = −0.61 between upper-arm length and the load that maximizes power suggests that a subject with a typically shorter upper-arm length (1 SD below the mean) produced maximal power at 38% 1RM; whereas a typically longer upper-arm subject (1 SD above the mean) produced maximal power at 22% 1RM. Although a very large correlation between bench press strength and estimated peak power was found (r = 0.76; 0.58 to 0.87), only a small unclear correlation between bench press strength and the percentage of 1RM that produced peak power was observed (r = −0.23; 0.10 to −0.52). Additionally, there was a small unclear correlation between absolute power and the percentage of 1RM that produced peak power (r = −0.23; 0.10 to −0.52).
The purpose of this investigation was to determine the extent in which anthropometric measures correlate with the load that produces maximal power using the methods outlined by Bevan et al. (9). Small to large correlations ranging from r = −0.29 (forearm length and upper-arm girth) to −0.61 (upper-arm length) were observed, and thus, the shared variance of these measures (r 2 as a %) suggest that 8–37% of the between-subject variation may be explained by these anthropometric descriptors of the upper limb. More specifically, players with typically long or short upper-arm length (1 SD from the mean) varied by ±8% 1RM in the load that produced maximal power. Whereas players with typically long or short forearm length or upper-arm girth varied by ±4% 1RM in the load that produced maximal power. These findings are of particular importance to practitioners when prescribing training loads to larger groups of players whereby greater within-athlete variation in arm segment lengths is likely.
The upper-arm and total arm length resulted in the strongest relationships between the percentage of 1RM that produced maximal power. The negative correlations associated with these measures suggest that subjects with longer upper-arm and total arm length produced their maximal power at a lower percentage of 1RM, with the opposite being true for those with shorter limb lengths. Interestingly, our results showed that athletes with larger upper-arm girth also had a negative correlation with the load that produced maximal power. We re-analyzed some of the data and noted that those with longer upper arms typically had larger girths (r = 0.42), and both upper-arm length and upper-arm girth were positively correlated with maximal strength (r = 0.34 and 0.57, respectively) and absolute power (r = 0.15 and 0.35, respectively). Thus, those with longer and bigger upper arms, as a generalization, were stronger and more powerful. Furthermore, our results showed that there were small, albeit unclear, negative relationships with the load that produces maximal power and absolute strength and power. These findings suggest that those with higher strength and absolute power produced maximal power at a lower percent of 1RM. These findings are also in line with previous findings by Baker (5) who reported that stronger athletes use a significantly lower percentage of 1RM to produce maximal power compared with their less strong counterparts. Therefore, the between-subject variation in the load that produces maximal power may be due to differences in (a combination of) anthropometric characteristics and absolute strength and power outputs.
Upper-body peak power occurred at 30% 1RM bench press and was similar to that reported by Bevan et al. (9) and Newton et al. (26) but lower than previously reported peak power outputs which occurred between 40 and 70% of 1RM bench press (5,6,12). It is likely that the difference in findings was because of methodologies used, which have been rigorously discussed elsewhere (2,5–7,9–11,16,17,20,25). One potential reason for the lower load observed in the present study is the use of assessing peak power rather than mean power. Previous authors have reported maximal mean power to be achieved at a 10–15% higher 1RM when compared with maximal peak power (12,25). Indeed, if maximal power was achieved at a 15% higher 1RM load in the current study (approximately 45% of 1RM), it would be consistent with previous literature that recruited a similar athletic population (professional Rugby League athletes; 40–60% 1RM) (4–6). It has been previously suggested that when reporting power-load relationships, peak power, rather than mean power, should be reported because of its larger relationship with athletic performance (14). Indeed, Marques et al. (23) reported that throwing velocity in team handball players was significantly correlated with peak bench throw power in 2 of 3 loads assessed; whereas, mean power was not significantly correlated with any loads measured. Additionally, Harman et al. (15) reported a correlation of r = 0.88 between vertical jump height and peak power but reported only a correlation of r = 0.54 between mean power and jump height.
When methods are replicated, the average percentage of 1RM strength that produced maximal power in a group of rugby players in the present study can be reproduced and is comparable with previous literature. However, relatively large individual variation is still observed and is in part because of between-subject differences in anthropometric characteristics and absolute strength and power outputs. Indeed, athletes with longer limbs and larger girths and greater maximal strength and power outputs utilize a lower percentage of 1RM loads to achieve maximum power. On the basis of these findings, it is recommended that all subjects be individually assessed to determine the load that produces maximal power.
The results of the present study do not constitute endorsement by the National Strength and Conditioning Association.
1. Argus CK, Gill N, Keogh J, Hopkins WG, Beaven CM. Effects of a short-term pre-season training programme on the body composition and anaerobic performance of professional rugby union players. J Sports Sci 28: 679–686, 2010.
2. Argus CK, Gill ND, Keogh JWL, Hopkins WG. Assessing lower-body peak power in elite rugby-union players. J Strength Cond Res 25: 1616–1621, 2011.
3. Argus CK, Gill ND, Keogh JWL, Hopkins WG, Beaven CM. Changes in strength, power and steroid hormones during a professional rugby union competition. J Strength Cond Res 23: 1583–1592, 2009.
4. Baker D. Comparison of maximum upper body strength and power between professional and college-aged rugby league football players. J Strength Cond Res 15: 30–35, 2001.
5. 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.
6. Baker D, Nance S, Moore M. The load that maximizes the average mechanical power output during explosive bench press throws in highly trained athletes. J Strength Cond Res 15: 20–24, 2001.
7. Baker D, Nance S, Moore M. The load that maximizes the average mechanical power output during jump squats in power-trained athletes. J Strength Cond Res 15: 92–97, 2001.
8. Batterham AM, Hopkins WG. Making meaningful inferences about magnitudes. Int J Sports Physiol Perf 1: 50–57, 2006.
9. Bevan HR, Bunce PJ, Owen NJ, Bennett MA, Cook CJ, Cunningham DJ, Newton RU, Kilduff LP. Optimal loading for the development of peak power output in professional rugby players. J Strength Cond Res 24: 43–47, 2010.
10. Cormie P, Grant OM, Triplett NT, McBride JM. Optimal loading for maximal power output during lower-body resistance exercises. Med Sci Sports Exerc 39: 340–349, 2007.
11. Cormie P, McBride JM, McCaulley GO. Power-time, force-time, and velocity-time curve analysis during the jump squat: Impact of load. J Appl Biomech 24: 112–120, 2008.
12. Cronin J, McNair PJ, Marshall RN. Developing explosive power: A comparison of technique and training. J Sci Med Sport 4: 59–70, 2001.
13. Drinkwater EJ, Galna B, McKenna MJ, Hunt PH, Pyne DB. Validation of an optical encoder during free weight resistance movements and analysis of bench press sticking point power during fatigue. J Strength Cond Res 21: 510–517, 2007.
14. Dugan EL, Doyle TLA, Humphries B, Hasson CJ, Newton RU. Determining the optimal load for jump squats: A review of methods and calculations. J Strength Cond Res 18: 668–674, 2004.
15. Harman EA, Rosenstein MT, Frykman PN, Rosenstein RM. The effects of arms and countermovement on vertical jumping. Med Sci Sports Exerc 22: 825–833, 1990.
16. Harris NK, Cronin JB, Hopkins WG. Power outputs of a machine squat-jump across a spectrum of loads. J Strength Cond Res 21: 1260–1264, 2007.
17. Harris NK, Cronin JB, Hopkins WG, Hansen KT. Squat jump training at maximal power loads vs. heavy loads: Effect on sprint ability. J Strength Cond Res 22: 1742–1749, 2008.
18. Hart CL, Ward TE, Mayhew JL. Anthropometric correlates of bench press performance following resistance training. Res Sports Med 2: 89–95, 1991.
19. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 41: 3–12, 2009.
20. Kaneko M, Fuchimoto T, Toji H, 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.
21. Lander J. Maximums based on reps. Natl Strength Cond Assoc J 6: 60–61, 1985.
22. LeSuer DA, McCormick JH, Mayhew JL, Wasserstein RL, Arnold MD. The accuracy of prediction equations for estimating 1-RM performance in the bench press, squat, and deadlift. J Strength Cond Res 11: 211–213, 1997.
23. Marques MC, van den Tillaar R, Vescovi JD, Gonzalez-Badillo JJ. Relationship between throwing velocity, muscle power, and bar velocity during bench press in elite handball players. Int J Sports Physiol Perf 2: 414, 2007.
24. Newton RU, Kraemer WJ, Häkkinen K, Humphries BJ, Murphy AJ. Kinematics, kinetics, and muscle activation during explosive upper body movements. J Appl Biomech 12: 31–43, 1996.
25. Newton RU, Murphy AJ, Humphries BJ, Kraemer WJ, Häkkinen K. Influence of load and stretch shortening cycle on the kinematics, kinetics and muscle activation that occurs during explosive upper body movements. Eur J Appl Physiol 75: 333–342, 1997.
26. Newton RU, Murphy AJ, Humphries BJ, Wilson GJ, Kraemer WJ, Häkkinen K. Influence of load and stretch shortening cycle on the kinematics, kinetics and muscle activation that occurs during explosive upper-body movements. Eur J Appl Physiol Occup Physiol 75: 333–342, 1997.
27. Norton K, Whittingham N, Carter L, Kerr D, Gore C, Marfell-Jones M. Measurement techniques in anthropometry. In: Anthropometrica. Norton K., Olds T., eds. Marrickville, Australia: UNSW Press, 1996. pp. 25–75.
28. Siegel JA, Gilders RM, Staron RS, Hagerman FC. Human muscle power output during upper- and lower-body exercises. J Strength Cond Res 16: 173–178, 2002.