One-Repetition Maximum Bench Press Performance Estimated With a New Accelerometer Method

SAVE OFFLINE

ONE-REPETITION MAXIMUM BENCH PRESS PERFORMANCE ESTIMATED WITH A NEW ACCELEROMETER METHOD JARI-PEKKA RONTU,1,2 MANNE I. HANNULA,1 SAMI LESKINEN,1 VESA LINNAMO,2 AND JUKKA A. SALMI2 1

School of Engineering, Oulu University of Applied Sciences, Oulu, Finland; and 2Neuromuscular Research Center, University of Jyva¨skyla¨, Jyva¨skyla¨, Finland

ABSTRACT Rontu, J-P, Hannula, MI, Leskinen, S, Linnamo, V, Salmi, JA. One-repetition maximum bench press performance estimated with a new accelerometer method. J Strength Cond Res 24(8): 2018–2025, 2010—The one repetition maximum (1RM) is an important method to measure muscular strength. The purpose of this study was to evaluate a new method to predict 1RM bench press performance from a submaximal lift. The developed method was evaluated by using different load levels (50, 60, 70, 80, and 90% of 1RM). The subjects were active floorball players (n = 22). The new method is based on the assumption that the estimation of 1RM can be calculated from the submaximal weight and the maximum acceleration of the submaximal weight during the lift. The submaximal bench press lift was recorded with a 3-axis accelerometer integrated to a wrist equipment and a data acquisition card. The maximum acceleration was calculated from the measurement data of the sensor and analyzed in personal computer with LabView-based software. The estimated 1RM results were compared with traditionally measured 1RM results of the subjects. An own estimation equation was developed for each load level, that is, 5 different estimation equations have been used based on the measured 1RM values of the subjects. The mean (6SD) of measured 1RM result was 69.86 (615.72) kg. The mean of estimated 1RM values were 69.85–69.97 kg. The correlations between measured and estimated 1RM results were high (0.89–0.97; p , 0.001). The differences between the methods were very small (20.11 to 0.01 kg) and were not significantly different from each other. The results of this study showed promising prediction accuracy for estimating bench press performance by performing just a single submaximal bench press lift. The estimation accuracy is competitive with other

Address correspondence to Jukka A. Salmi, [email protected]. 24(8)/2018–2025 Journal of Strength and Conditioning Research Ó 2010 National Strength and Conditioning Association

2018

the

known estimation methods, at least with the current study population.

KEY WORDS one repetition maximum, 1RM, submaximal weight, bench press, accelerometer INTRODUCTION

T

he one-repetition maximum (1RM) test is the most popularly used procedure to measure muscular strength (2,5–8,10,11,15). The term 1RM means the most weight that can be lifted successfully through a full range of motion used. The method is often used to measure upper-body muscular strength with basic lifts like bench press (5,9,14). Maximal tests of dynamic strength are commonly used for athletes, but for the elderly and physically inactive subjects, the tests may be difficult to perform. Even then, some study results indicate that with proper technique, the 1RM testing may be performed in, for example, cardiac rehabilitation patients without injury or significant muscle soreness or abnormal heart rate/rhythm or blood pressure responses (1). The 1RM testing has also been considered as dangerous because of dealing with heavy loads and impractical because the method is time consuming (2,6,7,10,14,15). The potential for injury may be increased with heavier loads, which cause a lot of stress on the muscles, bones, and connective tissues. The use of muscular endurance repetitions has been implemented to estimate the 1RM, for example, in the bench press (11,15). Several formulas have been developed to predict 1RM from submaximal repetitions to fatigue, and muscular endurance is normally highly correlated with 1RM result (3,5–7,10,11,14,15). Using higher loads leading to fewer repetitions may be more accurate method than higher repetition number with lighter loads. There is a strong correlation between 1RM and number of repetitions when 10 or fewer repetitions (3,5,10,11,14,15). The estimation results of 1RM with repetitions-to-fatigue (RTF) type of testing vary a lot depending on the gender and training history of the subject (4). Some studies have been made to establish a relationship between selected anthropometric dimensions and 1RM

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

TABLE 1. Characteristics of the subjects presented as mean (6SD). n = 22

Mean

Age (y) Weight (kg) Height (cm)

21.7 (64.5) 72.3 (65.6) 177.0 (65.5)

bench press performance (8,9,12). Using anthropometric dimensions, however, seem not to be a sufficient method, either alone or combined with other estimation methods, to get 1RM performance substantially better predicted. Many resistance training programs are based on relative loads derived from a person’s 1RM. An athlete may, for example, perform a specific number of sets and repetitions at different percentage levels of his/her 1RM. It is therefore important to have an accurate estimation method for 1RM because of the injury risk from over predicting 1RM. The testing of 1RM or muscular endurance testing is typically not feasible in the beginning of the training session with the normal training program execution. To avoid these drawbacks, a new method to estimate 1RM was developed. The method was based on acceleration analysis of the submaximal lift in bench press. A uniaxial accelerometer has previously been used to measure muscular power, and it has been shown that accelerometers can provide a reliable and

| www.nsca-jscr.org

versatile means to assess muscle power (13). To best of our knowledge, this would be the first attempt to use accelerometer-based estimation of maximum strength. The main advantage of the method developed would be the simple prediction of 1RM by only one execution of a submaximal lift. In addition, our aim was to develop a method with use of modern technology for quick performance interpretation. A training program could then easily be modified according to the present day’s performance level. The new method could be useful and simple enough in evaluating strength in a larger population of subjects. The purpose of this study was to examine if 1RM bench press performance could be reliably predicted with accelerometer from submaximal lift.

METHODS Experimental Approach to the Problem

In this study, we investigated the accuracy of the accelerometer-based novel method to estimate 1RM of bench press. We compared the new method against traditional 1RM test. We developed estimation equations to this new method for 50, 60, 70, 80, and 90% and one for a wider area of the bench press load to predict 1RM. The estimated value was then compared and tested with the absolute value defined by the traditional method. Subjects

The data were obtained from 22 Finnish male floorball players whose physical characteristics are shown in Table 1. Most of the subjects (19 players) played in Finnish

Figure 1. A typical acceleration signal during a bench press lift.

VOLUME 24 | NUMBER 8 | AUGUST 2010 |

2019

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Novel 1RM Prediction Method

Figure 2. The measurement setup.

championship league. They were training 3 times a week in an average, which was mainly sport specific in nature. The measurements were conducted at the end of the season. Before testing, each participant received an information description of the tests and signed a written informed consent statement. The chair of the university department, who acted as an institutional review member, approved this study project. All of the subjects were involved in active training programs, but the majority of the players were not familiar with bench press strength training.

Method

This experimental study was based on acceleration measurements conducted by a LIS3L02AQ 3-axis linear accelerometer (ST Microelecronics, Inc., Geneva, Switzerland). The accelerometer with the additional electronics was integrated to custom-built wrist equipment. The measurement signals were recorded with NI-6009 (National Instruments, Austin, TX, USA) data acquisition card using sampling rate 1 kHz. The total acceleration was calculated from the measurement ffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi data of each axis of the sensor (a tot ¼ a 2x þ a 2y þ a 2z ) and

TABLE 2. The means (6SD) and differences (6SD) of measured 1RM and estimated 1RM, 95% confidence intervals (CI) of the differences, and the correlations between the real 1RM and estimations. n = 22 Measured 1RM 50% load 60% load 70% load 80% load 90% load

Mean (kg) 69.864 69.855 69.864 69.973 69.855 69.864

Difference (kg)

(615.72) (613.96) (614.57) (615.39) (615.32) (615.27)

0.01 (67.23) 0.00 (65.91) 20.11 (64.34) 0.01 (63.55) 0.00 (63.75)

95% CI (kg) 23.20 to 22.62 to 22.03 to 21.57 to 21.66 to

3.21 2.62 1.82 1.58 1.66

Correlation 0.888* 0.927* 0.961* 0.974* 0.971*

*Significance value , 0.001.

2020

the

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

| www.nsca-jscr.org

the lift. It was assumed that with maximal effort, the force produced is Fmax = m
amax, where m is the submaximal weight used and amax is the maximal acceleration during the lift. If the same force is produced with 1RM (the acceleration of 9.81 m
s22), a basic prediction estimate of 1RM is

1RM ¼

a max
m: 9:81 m
s 2

The final estimation equations with different loads were defined from the real measurement data by fitting the line with the data:

1RM ¼ A
Figure 3. The correlation between the measured and estimated 1RM values.

analyzed in personal computer (PC) with LabView (National Instruments)-based software. The calibration of total acceleration was made by adjusting the acceleration value to 1 g (9.81 m
s22) when the bar was in a start position with the arms fully extended above the chest. A typical example of acceleration signal is shown in Figure 1. When the bar changes its direction from downward to upward, it causes a strong peak to the total acceleration signal. This peak was later removed from the analysis by ignoring a period of 220 milliseconds starting from the peak value. The wrist equipment was installed to the right wrist of the subject. Another wrist equipment was installed to the bar to get reference data during the same lift to clarify if soft tissues and joints will cause errors in measurements. The measurement setup is shown in Figure 2. The method is based on the assumption that the estimation of 1RM can be calculated from the submaximal weight and the maximum acceleration of the submaximal weight during

a max
mþB; 9:81 m
s 2

where A and B are coefficients of the fitted line equation. Testing Procedure

Normal standard Olympic bar and plates were used for the lifts. The bench press procedure was standard ‘‘touch-and-go’’ protocol (8,11). The subjects were assisted to get the bar from the support racks to a start position with the arms fully extended above the chest. The subject lowered the bar slowly to his chest and then returned the bar to the fully extended position. The weights were assisted back to the support racks. The subjects were in supine position on the bench during the execution. Subjects used a grip, which was a little bit wider (15–30 cm) than shoulder width. The test had to perform without chest bounce; foot movement; or head, upper back, and buttocks leaving the bench. The subjects were first allowed to make a proper warm-up as they wanted. Then the subjects made 2 bench press series of 10 lifts with 50% load of estimated 1RM result and finally one series of 4 repetitions with 60% load and one series of 4 repetitions with 80% load. The first test was a traditional 1RM test. Subjects were allowed to select the initial test weight based on previous training TABLE 3. Standard error of the estimate (SEE) and estimation errors at different load history. The target was to make levels. the first attempt with approxiLoad (n = 22) Standard error of the The mean absolute The mean absolute mately 90% load of estimated (%) estimate, SEE (kg) error (kg) error (%) 1RM. If the attempt was successful, additional weight was 50 7.1 5.8 8.4 added for next attempt. The 60 5.8 4.6 6.6 70 4.2 3.4 4.7 minimum increase of weight 80 3.5 2.6 4.1 was 3 kg. This procedure was 90 3.7 3.2 4.8 repeated until the subject failed to perform the lift. The highest successfully lifted weight was VOLUME 24 | NUMBER 8 | AUGUST 2010 |

2021

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Novel 1RM Prediction Method

Figure 4. The correlations between the measured and estimated 1RM values using different loads (approximately 50, 60, 70, 80, and 90% of measured 1RM).

recorded as the 1RM bench press result. The target was to find 1RM result within 3–5 attempts. The subjects had 5-minute rest interval between 1RM attempts. The subjects executed 5 single separate submaximal bench press lifts after the 1RM test. The rest interval between attempts was also 5 minutes. The used load levels were 50, 60, 70, 80, and 90% of the 1RM result measured in the first phase. The subjects were instructed to return the bar from chest to the fully extended position as fast as possible. The acceleration signals of both equipments (installed to the wrist and to the

2022

the

bar) were recorded during the lifts. The maximum acceleration values were defined from the data recorded, and the estimated 1RM results were calculated as described earlier. Statistical Analyses

The subject characteristics are presented as mean (6SD) and the measurement estimations as mean (6standard error of the estimate [SEE]), absolute error, and relative (%) error. The results were analyzed against measured 1RM results with Pearson correlation coefficient test and one-sample

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

| www.nsca-jscr.org

been used based on the measured 1RM values of the subTABLE 4. Standard error of the estimate (SEE) and the average estimation error when jects). The standard error of the a single equation was used (the equation was derived from the results of 70 and estimate (SEE) and the average 80% loads). estimation error (kg and %) of each load are presented in Load (n = 22) Standard error of the The mean absolute The mean absolute (%) estimate, SEE (kg) error (kg) error (%) Table 3. The data in Figure 3 collects the cases with different 50 19.1 17.4 24.2 loads into the same chart. 60 11.9 10.1 13.9 Significant correlations were 70 5.2 3.9 3.8 observed when the measured 80 4.5 3.2 4.6 90 12.8 12.1 17.7 1RM and estimated 1RM values were compared with each other. The correlations varied between 0.888 and 0.974. The (paired) t-test. The alpha level for statistical significance was graphs in Figure 4 present the correlations between the set to p # 0.05. measured 1RM and estimated 1RM using different loads. The developed estimation equations were

RESULTS

The average performance characteristics of the subjects are indicated in Table 2. The mean (6SD) of the measured 1RM result was 69.86 (615.72) kg. The mean of estimated 1RM values were 69.85–69.97 kg. The correlations between measured and estimated 1RM results were very good (0.89–0.97; p , 0.001). The differences between the methods were very small (20.11 to 0.01 kg) and were not significantly different (Table 2 and Figure 3). The measurements indicated that the estimation was improved with higher loads. The estimated 1RM values have been calculated separately with an own equation developed for each load level (i.e., 5 different estimation equations have

a max
m þ 7:8975 ð50%loadÞ 9:81 m
s 2 a max
mþ6:1938 ð60%loadÞ 1RM ¼ 1:2292
9:81 m
s 2 a max
mþ4:5297 ð70%loadÞ 1RM ¼ 1:1066
9:81 m
s 2 a max 1RM ¼ 1:0327

mþ3:2166 ð80%loadÞ 9:81 m
s 2 a max
mþ4:2391 ð90%loadÞ; 1RM ¼ 0:8918
9:81 m
s 2 1RM ¼ 1:3882


where amax is the maximal acceleration during the lift and m is the submaximal weight used. The main trend is that the slope of the fitted line decreases when the used load increases (from 1.3882 to 0.8918). The equations of 70 and 80% loads are quite close to each other. A common equation based on the results of 70 and 80% loads was also developed to cover a wider dynamic area of load.

a max 9:81 m
s 2
m þ 6:9836

1RM ¼ 1:0187


Figure 5. The correlation between the measured and estimated 1RM values when a single equation was used.

Standard error of the estimate and the average estimation error (kg and %) of each load are presented in Table 4. Standard error of the estimate was 12.0 kg, when data from all

VOLUME 24 | NUMBER 8 | AUGUST 2010 |

2023

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Novel 1RM Prediction Method load levels was used in the calculation. The correlation between the measured 1RM and estimated 1RM is presented in Figure 5. The correlation factor (n = 110) was 0.750 (p , 0.001). The correlation between the measured 1RM and estimated 1RM of 70 and 80% load results only, which had the smallest errors, was 0.949 (n = 44; p , 0.001). The correlation between the estimated 1RM results of different equipment (the first equipment was installed to the right wrist of the subject and another equipment was installed to the bar to get reference data during the same lift) was 0.973–0.991 (p , 0.001; n = 18). The correlations between correspondent maximum acceleration values were 0.607– 0.806 (p , 0.01; n = 18).

DISCUSSION Current submaximal tests can predict 1RM quite well, but they still take time or need many different measurements if anthropometric dimensions are used. The results of this study showed promising prediction accuracy for estimating bench press performance by performing just a single submaximal bench press lift. The purpose of this study was to evaluate the method with different loads. The estimation equations were developed individually for each load. The correlation coefficients varied between 0.88 and 0.97 and the SEE values between 3.5 and 7.1 kg. The best estimation accuracy was obtained with a load of 80% from measured 1RM (SEE was 3.5 kg and the mean absolute error was 2.6 kg and 4.1%). The SEE values of experienced weightlifters have been studied in many RTF type of tests. The best bench press equation in the study of Kravitz et al. (7) was obtained with 70% load with an SEE of 2.69 kg and correlation of 0.97. The subjects were experienced weightlifters. Mayhew et al. (10) determined the accuracy of predicting 1RM bench press performance in various groups of men (n = 220; untrained students, resistance-trained students, college wrestlers, soccer players, football players, high school students, and resistancetrained middle-aged men). The average 1RM was 96.6 kg. Six different relative endurance equations were used to predict the 1RM result. When subjects completed less than 10 repetitions, the SEE values varied between 4.0 and 4.5 kg and correlation was 0.98 in all of those cases. If subjects needed more than 10 repetitions, the SEE values were 6.8–16.7 kg and correlations 0.78–0.97. Mayhew et al. (11) made a study using college football players, who were experienced for heavy resistance training. The mean bench press 1RM was 136.7 kg. Seven prediction equations were validated, and 5 of those gave SEE values between 8.3 and 9.0 kg with correlations of 0.93–0.94 and the worst SEE values were 41.9 and 47.3 kg and correlations 0.43 and 0.49. In the study of Kim et al. (6), the subjects had limited weight training experience. The SEE values for men were 8.0 and 8.2 kg and for women, 2.7 and 3.1 kg. The corresponding correlations were 0.87 and 0.94 for men and 0.90 and 0.87 for women. The mean bench press 1RM for men was 88.9 kg and for women, 31.9 kg. Cummings and Finn (3) studied the estimation of

2024

the

1RM for untrained women using 3 different prediction equations. The SEE values were 1.753–1.778 kg and correlations 0.941–0.943. The mean of measured 1RM was 33.53 kg. Anthropometric dimensions have also been used to predict 1RM bench press performance (8,9,12). The correlation coefficients varied between 0.51 and 0.87 and the SEE values between 5.6 and 24.7 kg. Prediction of bench press strength from anthropometric dimensions seems to be less accurate than the other estimation methods. If SEE is proportioned with the average 1RM, the ratio SEE/1RM ratio 3 100% was 5.0% with a load of 80% of 1RM in our study. With RTF type of methods, the ratio has varied between 4.1 and 34.6% (3,6,7,10,11). The same ratio with anthropometric dimension–based methods varied between 9.5 and 21.0% (8,9,12). The best results of other methods have been achieved with experienced weightlifters. The subjects in our study were active floorball players. They were used to resistance exercises, but they had no real bench press training experience. The 1RM result might have uncertainty with novice lifters and cause bigger SEE values. Moreover, we cannot conclude the accuracy of the method for different size of athletes. It can, however, be concluded that the new method has a reasonable estimation accuracy and competitive with other known methods. The main problem in this new method is that the estimation equations developed are different for different loads. To increase the dynamic area of used load, a new common equation was developed. It gave promising results in the area of 70–80% load levels, but the estimation accuracy was substantially decreased when smaller or bigger load levels were used. Further studies should concentrate to develop a single estimation equation to cover a wider dynamic load level area. The significance of the placement of accelerometer was also investigated. The accelerometer was installed to the wrist of the subject and to the bar. The correlations between the estimated 1RM results of differently placed accelerometers were very good (0.97–0.99). It seems that the placement of the accelerometer and thus the soft tissues of the arm do not affect substantially to the 1RM bench press estimation results. The analyses were made manually by using the recorded measurement data. Some disturbances in measurement signals were detected despite using shielded measurement cables and signal filtering already in the wrist equipment. The disturbances were most probably caused by mains interruption in the laptop. One way to decrease the level of disturbances would be to use PC without mains current. The final application solution should be a stand-alone wrist equipment with an automatic real-time analysis program and a proper user interface with a display. This kind of solution could likely solve the disturbance problems. Those signal parts of the acceleration signal, which do not belong to the real lift itself (for instance, the signal part that is created when the bar changes direction on the chest of the subject), should be removed before the analysis. Calibration of acceleration values of the accelerometer needs also

TM

Journal of Strength and Conditioning Research

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

| www.nsca-jscr.org

more attention. Calibration process should also be done automatically in the beginning of each lift by software to give exact 9.81 m
s22 (1 g) when the equipment is in stable position. The new method based on just one submaximal lift gave promising bench press 1RM estimation results compared with other known methods. The main problem noticed in this new method was the small dynamic area for the estimation. The 1RM could not be estimated with just a single prediction equation but using several ones depending on the used percentual load level. The main target for development of the method is to develop a better estimation equation to cover a wider load level area and apply the method also in other muscular strength measurements.

2. Chapman, PP, Whitehead, JR, and Binkert, RH. The 225-lb reps-to-fatigue test as a submaximal estimate of 1-RM bench press performance in college football players. J Strength Cond Res 12: 258–261, 1998.

PRACTICAL APPLICATIONS

8. Mayhew, JL, Jacques, JA, Ware, JS, Chapman, PP, Bemben, MG, Ward, TE, and Slovak, JP. Anthropometric dimensions do not enhance one repetition maximum prediction from the NFL-225 test in college football players. J Strength Cond Res 18: 572–578, 2004.

The main benefit of the present method in addition to accuracy seems to be simpleness and flexibility. The 1RM can be estimated by lifting the submaximal weight only once and get the estimation immediately. This kind of method for maximum strength testing would be useful in many cases. Many resistance training programs are based on relative loads derived from a person’s 1RM. With the present method, it is possible to make the 1RM test even in the beginning of every training session if needed. A training program could be then easily modified according to the present day’s performance level. The method evaluated in this study was based on the use of accelerometer, which can be integrated either to specific wrist equipment or to existing equipment (e.g., wrist watches, sport monitors, mobile phones, and so on). Many heart rate monitors and mobile phones already have accelerometers inside. This makes the method interesting and widely applicable.

REFERENCES 1. Barnard, KL, Adams, KJ, Swank, AM, Mann, E, and Denny, DM. Injuries and muscle soreness during the one repetition maximum assessment in a cardiac rehabilitation population. J Cardiopulm Rehabil 19: 190–192, 1999.

3. Cummings, B and Finn, KJ. Estimation of a one repetition maximum bench press for untrained women. J Strength Cond Res 12: 262–265, 1998. 4. Fleck, SJ and Kraemer, WJ. Designing Resistance Training Programs (2nd ed.). Champaign, IL: Human Kinetics, 1997. 5. Horvat, M, Ramsey, V, Franklin, C, Gavin, C, Palumbo, T, and Glass, LA. A method for predicting maximal strength in collegiate women athletes. J Strength Cond Res 17: 324–328, 2003. 6. Kim, PS, Mayhew, JL, and Peterson, DF. A modified YMCA bench press test as a predictor of 1 repetition maximum bench press strength. J Strength Cond Res 16: 440–445, 2002. 7. Kravitz, L, Akalan, C, Nowicki, K, and Kinzey, SJ. Prediction of 1 repetition maximum in high-school power lifters. J Strength Cond Res 17: 167–172, 2003.

9. Mayhew, JL, Piper, FC, and Ware, JS. Anthropometric correlates with strength performance among resistance trained athletes. J Sports Med Phys Fitness 33: 159–165, 1993. 10. Mayhew, JL, Prinster, JL, Ware, JS, Zimmer, DL, Arabas, JR, and Bemben, MG. Muscular endurance repetitions to predict bench press strength in men of different training levels. J Sports Med Phys Fitness 35: 108–113, 1995. 11. Mayhew, JL, Ware, JS, Cannon, K, Corbett, S, Chapman, PP, Bemben, MG, Ward, TE, Farris, B, Juraszek, J, and Slovak, JP. Validation of the NFL-225 test for predicting 1-RM bench press performance in college football players. J Sports Med Phys Fitness 42: 304–308, 2002. 12. Scanlan, JM, Ballmann, KL, Mayhew, JL, and Lantz, CD. Anthropometric dimensions to predict 1-RM bench press in untrained females. J Sports Med Phys Fitness 39: 54–60, 1999. 13. Thompson, CJ and Bemben, MG. Reliability and comparability of the accelerometer as a measure of muscular power. Med Sci Sports Exerc 31: 897–902, 1999. 14. Ware, JS, Clemens, CT, Mayhew, JL, and Johnston, TJ. Muscular endurance repetitions to predict bench press and squat strength in college football players. J Strength Cond Res 9: 99–103, 1995. 15. Whisenant, MJ, Panton, LB, East, WB, and Broeder, CE. Validation of submaximal prediction equations for the 1 repetition maximum bench press test on a group of collegiate football players. J Strength Cond Res. 17: 221–227, 2003.

VOLUME 24 | NUMBER 8 | AUGUST 2010 |

2025

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.