RELIABILITY AND SMALLEST WORTHWHILE DIFFERENCE OF THE NFL-225 TEST IN NCAA DIVISION I FOOTBALL PLAYERS J. BRYAN MANN,1,2 PAT J. IVEY,1 WILLIAM F. BRECHUE,3
AND
JERRY L. MAYHEW4,5
Departments of 1Athletic Performance; and 2Physical Therapy, University of Missouri, Columbia, Missouri; 3 Department of Physical Education, Center for Physical Development Excellence, United States Military Academy, West Point, New York; 4Human Performance Laboratory, Truman State University, Kirksville, Missouri; and 5Department of Physiology, A. T. Still University of Health Sciences, Kirksville, Missouri ABSTRACT Mann, JB, Ivey, PJ, Brechue, WF, and Mayhew, JL. Reliability and smallest worthwhile difference of the NFL-225 test in NCAA Division I football players. J Strength Cond Res 28(5): 1427–1432, 2014—The NFL-225 test is widely used to assess the strength level and evaluate the progress of college football players during resistance training. Despite the studies evaluating the validity of this test, there are no reports assessing its reliability. The purpose of this study was to determine the reliability and smallest worthwhile difference (SWD) of the NFL225 test in Division I college football players. Seventy-two players were assessed for more than 3 weeks for the number of repetitions completed with a constant load of 102.3 kg (225 lbs) during winter conditioning. Test sessions occurred on the same day and at the same time 1 week apart. Intraclass correlation coefficients (ICCs) between weeks 1 and 2 (ICC = 0.987), weeks 2 and 3 (ICC = 0.981), and across weeks 1, 2, and 3 (ICC = 0.988) indicated high relative reliability. A small technical error (TE) (TE = 0.5 repetitions) provided strong absolute reliability. The SWD suggests that a change in performance of 3 repetitions or more after training would indicate a meaningful improvement in performance for this test.
KEY WORDS muscular endurance, bench press
strength,
absolute
muscular
INTRODUCTION
R
esistance training programs in college football have the major objectives of maximizing the strength and muscular endurance of each player. Because of the schedule imposed by the yearly football calendar, it may be difficult to have sustained resisAddress correspondence to Dr. Jerry L. Mayhew, [email protected]. 28(5)/1427–1432 Journal of Strength and Conditioning Research Ó 2014 National Strength and Conditioning Association
tance training cycles that last longer than 6–10 weeks. At the end of each of these training cycles, players may be tested to evaluate their degree of strength improvement. The limit in training time does not often allow the determination of reliability of performance tests. Such information would be vital to allow the differentiation between test familiarization and the true effect of training. At the upper end of the strength continuum, improvements may be quite small and approach the level of variation noted in repeated trials of a performance. Gains by players are often evaluated either in absolute or relative terms to judge the success of the training program with limited criteria for judging the meaningfulness of the change. One of the most common tests currently used in Division I football to assess upper-body strength is the NFL-225 test largely because of its acceptance by the National Football League (NFL) combine (21,22,24). Each player is required to perform as many bench press repetitions as possible without rest using a weight of 102.3 kg (225 lbs) regardless of his body mass or playing position. Although this test has been widely evaluated for its accuracy in estimating 1 repetition maximum (1RM) bench press (4,5,7,13–17,27), there has been no assessment of its reliability. Furthermore, the smallest worthwhile difference (SWD) for this test is unknown. Absolute reliability analysis would indicate the consistency with which players are able to repeat a performance on a given test over time (8). Relative reliability would provide the consistency with which players maintain their ranking within a group when performing a specific test (8). Smallest worthwhile difference is a value that takes into consideration typical test-to-test variation, thus allowing an estimate of the amount of change required to indicate a meaningful improvement because of training intervention. Knowledge of this value would provide a better understanding of test outcome and significance by allowing judgment of how much of a performance gain is due to real improvement and how much is due to testing familiarization. Finally, test-to-test variations will provide an estimate of measurement error that can be expected when conducting this test. VOLUME 28 | NUMBER 5 | MAY 2014 |
1427
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
NFL-225 Reliability It would be worthwhile for strength and conditioning specialists to have some idea of how much improvement would be considered a meaningful gain in NFL-225 performance resulting from training. Therefore, the purpose of this study was to determine the absolute and relative reliabilities and SWD of the NFL-225 test in major college football players.
METHODS Experimental Approach to the Problem
Many Division I college football programs now widely use the NFL-225 test to judge strength performance of players at various times during a yearly training cycle. Lacking is a good estimate of the reliability of this test in major college players. This study was designed to assess the absolute and relative reliabilities of the NFL-225 test and to determine the SWD to allow an assessment of the degree of meaningful improvement resulting from training. Players were tested each week for 3 consecutive weeks at the same time of the day during an off-season winter conditioning period for the maximum number of repetitions they could perform with 102.3 kg. Strong verbal encouragement was given at each test to maximize player participation. Subjects
The subjects (n = 72) for this study were members of a successful Division I program that was consistently ranked in the top 25 in the country and were measured at the conclusion of a 6-week off-season resistance training program. All players had a minimum of 5 years of heavy-resistance training, were skilled in the performance of the bench press maneuver, and had performed the NFL-225 test on several previous occasions. Only players who were free of any previous upper extremity injuries within the previous year were invited to participate. Players were divided into position groups designated as a big group (offensive and defensive tackles and offensive guards, n = 17), mid group (offensive backs, tightends, linebackers, defensive ends, and quarterbacks, n = 20), and skill group (wide receivers and defensive backs, n = 35). Subjects were also considered according to successful NFL-225 repetitions completed as group 1 (1–10 repetitions, n = 25), group 2 (11–19 repetitions, n = 35), and group 3 ($20 repetitions, n = 12). In addition, subjects were divided into 3 body mass groups based on the top, middle, and bottom one-third of the group: Group A # 90.9 kg (n = 22), group B = 91.0–109.1 kg (n = 26), and group C $ 109.1 kg (n = 24). Demographic and performance variables for the subjects by each of these groups are presented in Table 1. Participants were informed of the risks and benefits of the testing program and signed an informed consent document before testing. All testing protocols were approved by the university’s Institutional Review Board for studies involving human subjects. Procedures
The NFL-225 test was performed by having each player performed as many repetitions as possible without pause.
1428
the
Each player preceded each NFL-225 test by performing an individualized warm-up using various exercises. The warmup was concluded by performing several sets of progressively heavy bench presses, ranging from 50–80% of an estimated 1RM. After an adequate recovery, each player grasped the bar at their normal bench press grip, which was slightly outside the shoulder joint (15–25 cm). Although no mandatory cadence was imposed for the repetition tests, each player was encouraged to maintain a constant pace of their own choosing, with no more than a 2-second pause between each repetition. The bar was required to touch the chest on each repetition and was not allowed to bounce off it; the bar was then returned to full-arm extension. The head, upper back, and buttocks were maintained in contact with the bench throughout the test. The total number of repetitions completed to full-arm extension was recorded. The same procedure was repeated once per week for 3 consecutive weeks. Because this test was conducted as a major indicator for player performance evaluation, a high degree of verbal encouragement by players and coaches was offered during both occasions. Personal records were acknowledged by individual declarations and team salutes. Statistical Analyses
Separate paired t-tests were performed between each pair of trials, with Bonferroni correction for multiple t-tests, to determine systematic bias (8) and establish the 95% confidence intervals for the difference between successive trials. Typical error of measurement was computed using the forpffiffiffi mula: TE ¼ SDdiff = 2, where SDdiff represented the SD of the difference between 2 trials (8). Repeated-measures analysis of variance (ANOVA) over the 3 trials allowed the calculation of an average technical error (TE) from the square root of the mean square error (19). Typical percent error was represented by the group average of the ratio of the SDdiff for successive trials divided by the mean of successive trials for each subject (8). Smallest pffiffiffiworthwhile difference was calculated as 1.96 3 TE 3 2 (8). Intraclass correlation coefficient (ICC) was calculated according to the method detailed by Weir (26). The 95% Limits of Agreement was estimated by the method suggested by Bland and Altman (2). One-way ANOVA was used to assess differences among position groups, repetition groups, and body mass groups, with Bonferroni post hoc follow-up testing where significance was noted. The significance level was set at p # 0.05. Power exceeded 0.73 for all analyses.
RESULTS Performance data for each of the 3 NFL-225 test trials are given in Table 1 according to the various groupings. The big and mid groups performed significantly more repetitions than the skill group, while the heaviest weight group performed significantly more repetitions than the middleweight group which performed significantly more repetitions than
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.com
TABLE 1. Demographic and performance characteristics of the subjects by group.* Age (y) Position groups Big (n = 17) Mid (n = 20) Skill (n = 35) Rep groups 1–10 (n = 25) 11–19 (n = 35) .20 (n = 12) Weight groups A: ,90.9 kg (n = 22) B: 91.0–109.1 kg (n = 26) C: .109.1 kg (n = 24)
Height (cm)
Weight (kg)
Trial 1
Trial 2
Trial 3
ICC†
20.5 6 1.2 190.6 6 5.6z 130.8 6 10.5§ 17.4 6 8.3zk 18.4 6 7.5¶ 18.7 6 7.3 0.987 20.1 6 1.0 186.2 6 4.4 106.1 6 8.3 15.5 6 4.8 15.3 6 4.9 16.6 6 5.1¶ 0.986 20.2 6 1.0 182.3 6 6.6 90.0 6 5.6 9.9 6 4.9 10.3 6 4.8 11.9 6 5.4¶ 0.983 20.0 6 1.0 185.6 6 7.0 20.3 6 1.1 184.6 6 7.2 20.8 6 0.9 187.1 6 3.9
95.2 6 16.5# 6.4 6 3.0z** 7.1 6 3.2¶ 8.4 6 3.4¶ 0.955 104.4 6 15.3 14.5 6 2.6 14.9 6 2.8 16.1 6 3.1¶ 0.942 122.0 6 16.8 23.5 6 4.2 23.1 6 4.3 24.4 6 3.3 0.939
20.3 6 1.1 179.6 6 5.8†† 86.4 6 2.2†† 9.0 6 5.0zzz 9.4 6 5.1 11.3 6 5.8¶ 0.988 20.0 6 0.9 185.9 6 4.6 98.8 6 6.0 13.0 6 5.1 13.2 6 5.0 48.5 6 5.5¶ 0.985 20.5 6 1.1 190.1 6 5.3
126.1 6 11.6 17.3 6 7.0
17.8 6 6.5 18.3 6 6.3 0.984
*ICC = intraclass correlation coefficient. †ICC values were calculated across all 3 trials. zBig = mid , skill. §Big . mid . skill. kTrial 1 = trial 2 , trial 3. ¶Group 1 = group 2 , group 3. #Group 1 , group 2 , group 3. **Group A , group B , group C. ††Group A = group B , group C. zzTrail significantly greater than previous trial.
TABLE 2. Absolute and relative reliabilities for NFL-225 test in Division I college football players (n = 72).
Constant error* %Change TEz CV%§ ICCk SEM¶ SWD# SWD%** LoA††
Trial 1 vs. 2
Trial 2 vs. 3
0.4 6 1.5† 7.5 6 24.2 1.0 7.7 0.987 0.2 2.9 21.8 22.1 to 3.7
1.2 6 1.8† 14.6 6 22.0 1.3 8.9 0.981 0.2 3.5 25.7 22.3 to 4.7
*Constant error (CE) calculated as the difference between successive sessions. †Significantly different between pffiffiffi trials (p # 0.05). zTypical error, TE = SDdiff/ 2. §CV% is the coefficient of variation calculated from the SDdiff for successive trials divided by the mean of successive trials for each subject. kIntraclass correlation coefficient (ICC) between successive sessions. pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¶Standard error of measurement, SDdiff 12ICC. #Smallest worthwhile difference, SWD = 1.96 3 pffiffiffi TE 3 2. **SWD% = SWD/mean (trial 1 and trial 2). ††LoA = 95% limits of agreement using the BlandAltman method (2).
the lightest group. Because the categorizations by position and weight groups followed similar distributions, an analysis of covariance holding body mass constant revealed no significant difference among the adjusted means (6SE) for the big group (12.3 6 2.6 repetitions), mid group (15.3 6 1.2 repetitions), and skill group (13.8 6 1.5 repetitions). Players made small but significant increases from trial 1 to trial 2 and from trial 2 to trial 3 when the entire group was considered (Table 2). When players were considered by the various group breakdowns, the big n group and the 1–10 repetition group tended to increase significantly from trial 1 to trial 2 (1 and 0.7 repetitions, respectively), whereas the mid and skill groups (1.3 and 1.6 repetitions, respectively) and the 1–10 and 11–19 repetition groups (1.2 and 1.3 repetitions, respectively) made significant increases from trial 2 to trial 3. All data are summarized by group in Table 1. Positive and negative changes in repetitions by players produced large variability in the percent change between each trial pair (Table 2), making this determination less desirable for reporting improvement. Intraclass correlation coefficients for the various groups are shown in Table 1. For the entire sample, ICCs between successive trials indicated strong performance consistency (relative reliability) with statistical power exceeding 0.99 (Table 2). Typical error and typical percent error (CV%) were acceptable and produced a SWD of 3 repetition. Thus, an increase of 3 repetitions or more in the NFL-225 test would indicate a meaningful gain VOLUME 28 | NUMBER 5 | MAY 2014 |
1429
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
NFL-225 Reliability in test outcome. The Bland-Altman plots confirmed a 95% confidence interval of 3 repetitions for the difference between trials 1 and 2 and 4 repetitions for the difference between trials 2 and 3 (Figures 1A, B). Hopkins has suggested that regression of 1 trial on another will reveal any bias that may exist (9). As noted in Figures 2A, B, the slopes of the regression lines between trials were near 1.00 and the intercept was approximately 1.0, indicating no systematic bias between the trials (8). Rhea (20) further supports the use of effect size (ES) to assess the magnitude of a difference between 2 means. The ES between trials 1 and 2 (ES = 0.03) and between trials 2 and 3 (ES = 0.09) indicated no practical difference between either pair of trials.
DISCUSSION
Figure 1. A, Bland-Altman plot illustrating level of agreement between trial 1 and trial 2. B, Bland-Altman plot illustrating level of agreement between trial 2 and trial 3.
Figure 2. A, Regression of trail 2 on trial 1 illustrating the lack of bias between trials. B, Regression of trial 3 on trial 2 illustrating the lack of bias between trials.
1430
the
Previous studies have noted a high correlation between NFL-225 repetitions and 1RM (4,5,7,13–17,27). However, this is the first study to analyze absolute and relative reliability for the NFL-225 test in a large sample of elite college players. According to the present data and the criteria given by Hopkins (8), the consistency of the NFL-225 test is very high and indicates that players should be able to repeat their performance within the same week with a variation of 62 repetitions. The major finding of this study suggests that improvements of 3 or more repetitions in the NFL-225 test would be the smallest change indicative of a meaningful effect. This should provide a guideline for determining the impact of various resistance training programs on NFL-225 performance in college players. A major dilemma in using the NFL-225 test as the determinant of improvement in strength resulting from a resistance training program may center on the nature of the test itself. Because a constant load is used by each player, the test is a measure of absolute muscular endurance (25). Absolute load endurance tests typically result in a greater numbers of repetitions being produced by larger or stronger individuals (12,25; Table 1), which could suggest the test is more an evaluation of muscle endurance than an assessment of strength. The number of repetitions performed by the heavier players places the test more in the endurancestrength zone of the performance continuum while being in the strength-endurance zone for smaller players (12,28). Based on an equation developed in a previous study (13), players in this study weighing more than 100.5 kg (median weight) were estimated to perform their NFL-225 repetitions at 44.6% (621.2%) of 1RM and produce approximately 17.3 (65.9) repetitions, whereas those less than 100.5 kg would perform at 67.9% (615.2%) of 1RM and produce approximately 10.4 (65.9) repetitions. In an earlier study on Division II players (23), the correlation between body weight and repetitions performed at 110% of body weight was nonsignificant (r = 0.17). Furthermore, there was no significant difference between light (,91.0 kg) and heavy (.91.0 kg) players in volume load (load 3 repetitions) estimated using a RepWt of 110% of
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 body weight. A more recent study of Division II players produced similar results (3), noting a nonsignificant correlation between body weight and repetitions at 70% (r = 20.19), 80% (r = 20.11), and 90% of 1RM (r = 20.16). In addition, volume load did not change significantly at each %1RM after a resistance training program that emphasized heavy loads and few repetitions (3). There was also no significant difference between weaker (,113.6 kg) and stronger (.136.4 kg) players in repetitions at each %1RM. However, stronger players achieved significantly greater volume loads at each %1RM than weaker players because of the greater RepWt for each %1RM. This contradicted earlier work on average subjects where the weaker individuals appeared to have an advantage in work performed (1). Although mass has be shown to be significantly related to the NFL-225 test and 1RM (13), statistically controlling for body weight reduced the correlation between 1RM and NFL-225 repetitions (r = 0.96) only slightly (r = 0.94). Therefore, as suggested by Naclerio et al. (18) and from the present observations, the goal of a training program to increase an absolute endurance task, such as the NFL-225 test, should be to maximize a player’s absolute strength level or perhaps his body mass. Because of the use of the NFL-225 test as the exclusive measure of upper-body strength performance at the NFL combine (21,22,24), players preparing to perform well in that event often alter their training to produce greater endurance in the bench press than strength improvement. Indeed, anecdotal evidence from various fitness organizations specializing in preparing players for NFL combine performance suggests that a greater emphasis on high repetitions in the bench press, in conjunction with loaded and unloaded push-ups, should be used. A limited amount of research has focused on high-repetition, low-load training as a means of strength improvement (1,6,10). Given the overwhelming amount of research supporting heavy-load, low-repetition training for strength improvement and the strong connection between absolute strength and repetition performance, it is unlikely that major college strength and conditioning programs should alter their resistance training programs to enhance NFL-225 performance without strong support for the effect of such a change on upper-body strength.
PRACTICAL APPLICATIONS The NFL-225 test is widely used in college football under the supposition of measuring muscular strength in the upper body. Previous studies have observed high correlations between 1RM bench press and the number of repetitions completed in the NFL-225 test while noting an increasing prediction bias at the upper end of the repetition continuum (14–17). This study found that the NFL-225 test is highly reproducible leading to the identification of a smallest worthwhile change of 3 repetitions to denote a significant change in performance greater than between-day variability (1–2 repetitions). Throughout an athlete’s career, it is likely
| www.nsca.com
that increases in absolute or relative strength might be one of the most beneficial factors for improving sports performance (11,12,28). However, considering the importance of showing improvement during a player’s college career and owing to the impact of performance scores at the NFL combine on professional draft status (24), one must consider the degree to which training specifically for this test becomes a priority.
REFERENCES 1. Anderson, T and Kearney, JT. Effect of three resistacne training programs on muscular strength and absolute and relative endurance. Res Q Exerc Sport 53: 1–7, 1982. 2. Bland, JM and Altman, DG. Statistical method of assessing agreement between two methods of clinical measurement. Lancet 327: 307–310, 1986. 3. Brechue, WF and Mayhew, JL. Upper-body work capacity and 1RM prediction are unaltered by increasing muscular strength in college football players. J Strength Cond Res 23: 2477–2486, 2009. 4. Brown, BR, Williams, T, Randolph, KL, and Mayhew, JL. Relationship of anthropometric dimensions to 1-RM bench press and NFL-225 test in college football players. Mo J Hlth Phys Educ Rec Dance 23: 36–45, 2013. 5. Chapman, PP, Whitehead, JR, and Binkert, RH. The 225-lb reps-to-fatigue test as a submaximal estimate of 1RM bench press performance in college football players. J Strength Cond Res 12: 258– 261, 1998. 6. Goto, K, Nagasawa, M, Yanagisawa, O, Kizuka, T, Ishii, N, and Takamatsu, K. Muscular adaptations to combinations of high- and low-intensity resistance exercise. J Strength Cond Res 18: 730–737, 2004. 7. Hetzler, RK, Schroeder, BL, Wages, JJ, Stickley, CD, and Kimura, IF. Anthropometry increases 1 repetition maximum predictive ability of NFL-225 test for division IA college football players. J Strength Cond Res 24: 1429–1439, 2010. 8. Hopkins, WG. Measures of reliability in sports medicine and science. Sports Med 30: 1–15, 2000. 9. Hopkins, WG. Bias in Bland-Altman but not regression validity analyses. Sportscience 8: 42–46, 2004. 10. Jackson, NP, Hickey, MS, and Reiser, RF II. High resistance/low repetition vs. low resistance/high repetition training: Effects on performance of trained cyclists. J Strength Cond Res 21: 289–295, 2007. 11. Kraemer, WJ and Fleck, SJ. Optimizing Strength Training: Designing Nonlinear Periodization Workouts. Champaign, IL: Human Kinetics, 2007. 12. Kurz, T. Science of Sports Training. Island Pond, VT: Stadion Publishing, 1991. 13. Mann, JB, Stoner, JB, and Mayhew, JL. NFL-225 test to predict 1RM bench press in NCAA division I football players. J Strength Cond Res 26: 2623–2631, 2012. 14. Mayhew, JL, Jacques, JA, Ware, JA, Chapman, PP, Bemben, MG, Ward, TE, and Slovak, JP. Anthropometric dimensions do not enhance 1-RM prediction from the NFL-225 test in college football players. J Strength Cond Res 18: 572–578, 2004. 15. Mayhew, JL, Srnka, S, Getty, C, Ball, A, and Jacques, JA. Evaluation of the NFL-225 test for predicting one repetition maximum bench press in small-college football players. Mo J Hlth Phys Educ Rec Dance 20: 67–76, 2010. 16. Mayhew, JL, Ware, JS, Bemben, MG, Wilt, B, Ward, TE, Farris, B, Juraszek, J, and Slovak, JP. The NFL-225 test as a measure of bench press strength in college football players. J Strength Cond Res 13: 130–134, 1999. VOLUME 28 | NUMBER 5 | MAY 2014 |
1431
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
NFL-225 Reliability 17. 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.
23. Schell, J, Ware, JS, and Mayhew, JL. Muscular endurance performance relative to body weight to predict 1-RM bench press in college football players. Iowa Assoc Hlth Phys Educ Rec Dance J 32: 30–31, 1999.
18. Naclerio, FJ, Colado, JC, Rhea, MR, Bunker, D, and Triplett, NT. The influence of strength and power on muscle endurance test performance. J Strength Cond Res 23: 1482–1488, 2009.
24. Sierer, SP, Battaglini, CL, Mihalik, JP, Shields, EW, and Tomasini, NT. The National Football combine: Performance differences between drafted and nondrafted players entering the 2004 and 2005 drafts. J Strength Cond Res 22: 6–12, 2008.
19. Phillips, WT, Batterham, AM, Valenzuela, JE, and Burkett, LN. Reliability of maximal strength testing in older adults. Arch Phys Med Rehabil 85: 329–334, 2004.
25. Stone, MH, Sands, WA, Pierce, KC, Newton, RU, Haff, GG, and Carlock, J. Maximum strength and strength training—A relationship to endurance? Strength Cond J 28: 44–53, 2006.
20. Rhea, MR. Determining the magnitude of treatment effects in strength training research through the use of the effect size. J Strength Cond Res 18: 918–920, 2004.
26. Weir, JP. Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res 19: 231– 240, 2005.
21. Robbins, DW. Positional characteristics of players drafted into the National Football League. J Strength Cond Res 25: 2661–2667, 2011.
27. Whisenant, MJ, Panton, LB, East, WB, and Broeder, CE. Validation of submaximal prediction equations for the 1 repetition maximum bench press on a group of collegiate football players. J Strength Cond Res 17: 221–227, 2003.
22. Robbins, DW. Relationships between national football league combine performance measures. J Strength Cond Res 26: 226–231, 2012.
1432
the
28. Zatsiorsky, VM and Kraemer, WJ. Science and Practice of Strength Training. Champaign, IL: Human Kinetics, 2006.
TM
Journal of Strength and Conditioning Research
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
AND
JERRY L. MAYHEW4,5
Departments of 1Athletic Performance; and 2Physical Therapy, University of Missouri, Columbia, Missouri; 3 Department of Physical Education, Center for Physical Development Excellence, United States Military Academy, West Point, New York; 4Human Performance Laboratory, Truman State University, Kirksville, Missouri; and 5Department of Physiology, A. T. Still University of Health Sciences, Kirksville, Missouri ABSTRACT Mann, JB, Ivey, PJ, Brechue, WF, and Mayhew, JL. Reliability and smallest worthwhile difference of the NFL-225 test in NCAA Division I football players. J Strength Cond Res 28(5): 1427–1432, 2014—The NFL-225 test is widely used to assess the strength level and evaluate the progress of college football players during resistance training. Despite the studies evaluating the validity of this test, there are no reports assessing its reliability. The purpose of this study was to determine the reliability and smallest worthwhile difference (SWD) of the NFL225 test in Division I college football players. Seventy-two players were assessed for more than 3 weeks for the number of repetitions completed with a constant load of 102.3 kg (225 lbs) during winter conditioning. Test sessions occurred on the same day and at the same time 1 week apart. Intraclass correlation coefficients (ICCs) between weeks 1 and 2 (ICC = 0.987), weeks 2 and 3 (ICC = 0.981), and across weeks 1, 2, and 3 (ICC = 0.988) indicated high relative reliability. A small technical error (TE) (TE = 0.5 repetitions) provided strong absolute reliability. The SWD suggests that a change in performance of 3 repetitions or more after training would indicate a meaningful improvement in performance for this test.
KEY WORDS muscular endurance, bench press
strength,
absolute
muscular
INTRODUCTION
R
esistance training programs in college football have the major objectives of maximizing the strength and muscular endurance of each player. Because of the schedule imposed by the yearly football calendar, it may be difficult to have sustained resisAddress correspondence to Dr. Jerry L. Mayhew, [email protected]. 28(5)/1427–1432 Journal of Strength and Conditioning Research Ó 2014 National Strength and Conditioning Association
tance training cycles that last longer than 6–10 weeks. At the end of each of these training cycles, players may be tested to evaluate their degree of strength improvement. The limit in training time does not often allow the determination of reliability of performance tests. Such information would be vital to allow the differentiation between test familiarization and the true effect of training. At the upper end of the strength continuum, improvements may be quite small and approach the level of variation noted in repeated trials of a performance. Gains by players are often evaluated either in absolute or relative terms to judge the success of the training program with limited criteria for judging the meaningfulness of the change. One of the most common tests currently used in Division I football to assess upper-body strength is the NFL-225 test largely because of its acceptance by the National Football League (NFL) combine (21,22,24). Each player is required to perform as many bench press repetitions as possible without rest using a weight of 102.3 kg (225 lbs) regardless of his body mass or playing position. Although this test has been widely evaluated for its accuracy in estimating 1 repetition maximum (1RM) bench press (4,5,7,13–17,27), there has been no assessment of its reliability. Furthermore, the smallest worthwhile difference (SWD) for this test is unknown. Absolute reliability analysis would indicate the consistency with which players are able to repeat a performance on a given test over time (8). Relative reliability would provide the consistency with which players maintain their ranking within a group when performing a specific test (8). Smallest worthwhile difference is a value that takes into consideration typical test-to-test variation, thus allowing an estimate of the amount of change required to indicate a meaningful improvement because of training intervention. Knowledge of this value would provide a better understanding of test outcome and significance by allowing judgment of how much of a performance gain is due to real improvement and how much is due to testing familiarization. Finally, test-to-test variations will provide an estimate of measurement error that can be expected when conducting this test. VOLUME 28 | NUMBER 5 | MAY 2014 |
1427
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
NFL-225 Reliability It would be worthwhile for strength and conditioning specialists to have some idea of how much improvement would be considered a meaningful gain in NFL-225 performance resulting from training. Therefore, the purpose of this study was to determine the absolute and relative reliabilities and SWD of the NFL-225 test in major college football players.
METHODS Experimental Approach to the Problem
Many Division I college football programs now widely use the NFL-225 test to judge strength performance of players at various times during a yearly training cycle. Lacking is a good estimate of the reliability of this test in major college players. This study was designed to assess the absolute and relative reliabilities of the NFL-225 test and to determine the SWD to allow an assessment of the degree of meaningful improvement resulting from training. Players were tested each week for 3 consecutive weeks at the same time of the day during an off-season winter conditioning period for the maximum number of repetitions they could perform with 102.3 kg. Strong verbal encouragement was given at each test to maximize player participation. Subjects
The subjects (n = 72) for this study were members of a successful Division I program that was consistently ranked in the top 25 in the country and were measured at the conclusion of a 6-week off-season resistance training program. All players had a minimum of 5 years of heavy-resistance training, were skilled in the performance of the bench press maneuver, and had performed the NFL-225 test on several previous occasions. Only players who were free of any previous upper extremity injuries within the previous year were invited to participate. Players were divided into position groups designated as a big group (offensive and defensive tackles and offensive guards, n = 17), mid group (offensive backs, tightends, linebackers, defensive ends, and quarterbacks, n = 20), and skill group (wide receivers and defensive backs, n = 35). Subjects were also considered according to successful NFL-225 repetitions completed as group 1 (1–10 repetitions, n = 25), group 2 (11–19 repetitions, n = 35), and group 3 ($20 repetitions, n = 12). In addition, subjects were divided into 3 body mass groups based on the top, middle, and bottom one-third of the group: Group A # 90.9 kg (n = 22), group B = 91.0–109.1 kg (n = 26), and group C $ 109.1 kg (n = 24). Demographic and performance variables for the subjects by each of these groups are presented in Table 1. Participants were informed of the risks and benefits of the testing program and signed an informed consent document before testing. All testing protocols were approved by the university’s Institutional Review Board for studies involving human subjects. Procedures
The NFL-225 test was performed by having each player performed as many repetitions as possible without pause.
1428
the
Each player preceded each NFL-225 test by performing an individualized warm-up using various exercises. The warmup was concluded by performing several sets of progressively heavy bench presses, ranging from 50–80% of an estimated 1RM. After an adequate recovery, each player grasped the bar at their normal bench press grip, which was slightly outside the shoulder joint (15–25 cm). Although no mandatory cadence was imposed for the repetition tests, each player was encouraged to maintain a constant pace of their own choosing, with no more than a 2-second pause between each repetition. The bar was required to touch the chest on each repetition and was not allowed to bounce off it; the bar was then returned to full-arm extension. The head, upper back, and buttocks were maintained in contact with the bench throughout the test. The total number of repetitions completed to full-arm extension was recorded. The same procedure was repeated once per week for 3 consecutive weeks. Because this test was conducted as a major indicator for player performance evaluation, a high degree of verbal encouragement by players and coaches was offered during both occasions. Personal records were acknowledged by individual declarations and team salutes. Statistical Analyses
Separate paired t-tests were performed between each pair of trials, with Bonferroni correction for multiple t-tests, to determine systematic bias (8) and establish the 95% confidence intervals for the difference between successive trials. Typical error of measurement was computed using the forpffiffiffi mula: TE ¼ SDdiff = 2, where SDdiff represented the SD of the difference between 2 trials (8). Repeated-measures analysis of variance (ANOVA) over the 3 trials allowed the calculation of an average technical error (TE) from the square root of the mean square error (19). Typical percent error was represented by the group average of the ratio of the SDdiff for successive trials divided by the mean of successive trials for each subject (8). Smallest pffiffiffiworthwhile difference was calculated as 1.96 3 TE 3 2 (8). Intraclass correlation coefficient (ICC) was calculated according to the method detailed by Weir (26). The 95% Limits of Agreement was estimated by the method suggested by Bland and Altman (2). One-way ANOVA was used to assess differences among position groups, repetition groups, and body mass groups, with Bonferroni post hoc follow-up testing where significance was noted. The significance level was set at p # 0.05. Power exceeded 0.73 for all analyses.
RESULTS Performance data for each of the 3 NFL-225 test trials are given in Table 1 according to the various groupings. The big and mid groups performed significantly more repetitions than the skill group, while the heaviest weight group performed significantly more repetitions than the middleweight group which performed significantly more repetitions than
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.com
TABLE 1. Demographic and performance characteristics of the subjects by group.* Age (y) Position groups Big (n = 17) Mid (n = 20) Skill (n = 35) Rep groups 1–10 (n = 25) 11–19 (n = 35) .20 (n = 12) Weight groups A: ,90.9 kg (n = 22) B: 91.0–109.1 kg (n = 26) C: .109.1 kg (n = 24)
Height (cm)
Weight (kg)
Trial 1
Trial 2
Trial 3
ICC†
20.5 6 1.2 190.6 6 5.6z 130.8 6 10.5§ 17.4 6 8.3zk 18.4 6 7.5¶ 18.7 6 7.3 0.987 20.1 6 1.0 186.2 6 4.4 106.1 6 8.3 15.5 6 4.8 15.3 6 4.9 16.6 6 5.1¶ 0.986 20.2 6 1.0 182.3 6 6.6 90.0 6 5.6 9.9 6 4.9 10.3 6 4.8 11.9 6 5.4¶ 0.983 20.0 6 1.0 185.6 6 7.0 20.3 6 1.1 184.6 6 7.2 20.8 6 0.9 187.1 6 3.9
95.2 6 16.5# 6.4 6 3.0z** 7.1 6 3.2¶ 8.4 6 3.4¶ 0.955 104.4 6 15.3 14.5 6 2.6 14.9 6 2.8 16.1 6 3.1¶ 0.942 122.0 6 16.8 23.5 6 4.2 23.1 6 4.3 24.4 6 3.3 0.939
20.3 6 1.1 179.6 6 5.8†† 86.4 6 2.2†† 9.0 6 5.0zzz 9.4 6 5.1 11.3 6 5.8¶ 0.988 20.0 6 0.9 185.9 6 4.6 98.8 6 6.0 13.0 6 5.1 13.2 6 5.0 48.5 6 5.5¶ 0.985 20.5 6 1.1 190.1 6 5.3
126.1 6 11.6 17.3 6 7.0
17.8 6 6.5 18.3 6 6.3 0.984
*ICC = intraclass correlation coefficient. †ICC values were calculated across all 3 trials. zBig = mid , skill. §Big . mid . skill. kTrial 1 = trial 2 , trial 3. ¶Group 1 = group 2 , group 3. #Group 1 , group 2 , group 3. **Group A , group B , group C. ††Group A = group B , group C. zzTrail significantly greater than previous trial.
TABLE 2. Absolute and relative reliabilities for NFL-225 test in Division I college football players (n = 72).
Constant error* %Change TEz CV%§ ICCk SEM¶ SWD# SWD%** LoA††
Trial 1 vs. 2
Trial 2 vs. 3
0.4 6 1.5† 7.5 6 24.2 1.0 7.7 0.987 0.2 2.9 21.8 22.1 to 3.7
1.2 6 1.8† 14.6 6 22.0 1.3 8.9 0.981 0.2 3.5 25.7 22.3 to 4.7
*Constant error (CE) calculated as the difference between successive sessions. †Significantly different between pffiffiffi trials (p # 0.05). zTypical error, TE = SDdiff/ 2. §CV% is the coefficient of variation calculated from the SDdiff for successive trials divided by the mean of successive trials for each subject. kIntraclass correlation coefficient (ICC) between successive sessions. pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¶Standard error of measurement, SDdiff 12ICC. #Smallest worthwhile difference, SWD = 1.96 3 pffiffiffi TE 3 2. **SWD% = SWD/mean (trial 1 and trial 2). ††LoA = 95% limits of agreement using the BlandAltman method (2).
the lightest group. Because the categorizations by position and weight groups followed similar distributions, an analysis of covariance holding body mass constant revealed no significant difference among the adjusted means (6SE) for the big group (12.3 6 2.6 repetitions), mid group (15.3 6 1.2 repetitions), and skill group (13.8 6 1.5 repetitions). Players made small but significant increases from trial 1 to trial 2 and from trial 2 to trial 3 when the entire group was considered (Table 2). When players were considered by the various group breakdowns, the big n group and the 1–10 repetition group tended to increase significantly from trial 1 to trial 2 (1 and 0.7 repetitions, respectively), whereas the mid and skill groups (1.3 and 1.6 repetitions, respectively) and the 1–10 and 11–19 repetition groups (1.2 and 1.3 repetitions, respectively) made significant increases from trial 2 to trial 3. All data are summarized by group in Table 1. Positive and negative changes in repetitions by players produced large variability in the percent change between each trial pair (Table 2), making this determination less desirable for reporting improvement. Intraclass correlation coefficients for the various groups are shown in Table 1. For the entire sample, ICCs between successive trials indicated strong performance consistency (relative reliability) with statistical power exceeding 0.99 (Table 2). Typical error and typical percent error (CV%) were acceptable and produced a SWD of 3 repetition. Thus, an increase of 3 repetitions or more in the NFL-225 test would indicate a meaningful gain VOLUME 28 | NUMBER 5 | MAY 2014 |
1429
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
NFL-225 Reliability in test outcome. The Bland-Altman plots confirmed a 95% confidence interval of 3 repetitions for the difference between trials 1 and 2 and 4 repetitions for the difference between trials 2 and 3 (Figures 1A, B). Hopkins has suggested that regression of 1 trial on another will reveal any bias that may exist (9). As noted in Figures 2A, B, the slopes of the regression lines between trials were near 1.00 and the intercept was approximately 1.0, indicating no systematic bias between the trials (8). Rhea (20) further supports the use of effect size (ES) to assess the magnitude of a difference between 2 means. The ES between trials 1 and 2 (ES = 0.03) and between trials 2 and 3 (ES = 0.09) indicated no practical difference between either pair of trials.
DISCUSSION
Figure 1. A, Bland-Altman plot illustrating level of agreement between trial 1 and trial 2. B, Bland-Altman plot illustrating level of agreement between trial 2 and trial 3.
Figure 2. A, Regression of trail 2 on trial 1 illustrating the lack of bias between trials. B, Regression of trial 3 on trial 2 illustrating the lack of bias between trials.
1430
the
Previous studies have noted a high correlation between NFL-225 repetitions and 1RM (4,5,7,13–17,27). However, this is the first study to analyze absolute and relative reliability for the NFL-225 test in a large sample of elite college players. According to the present data and the criteria given by Hopkins (8), the consistency of the NFL-225 test is very high and indicates that players should be able to repeat their performance within the same week with a variation of 62 repetitions. The major finding of this study suggests that improvements of 3 or more repetitions in the NFL-225 test would be the smallest change indicative of a meaningful effect. This should provide a guideline for determining the impact of various resistance training programs on NFL-225 performance in college players. A major dilemma in using the NFL-225 test as the determinant of improvement in strength resulting from a resistance training program may center on the nature of the test itself. Because a constant load is used by each player, the test is a measure of absolute muscular endurance (25). Absolute load endurance tests typically result in a greater numbers of repetitions being produced by larger or stronger individuals (12,25; Table 1), which could suggest the test is more an evaluation of muscle endurance than an assessment of strength. The number of repetitions performed by the heavier players places the test more in the endurancestrength zone of the performance continuum while being in the strength-endurance zone for smaller players (12,28). Based on an equation developed in a previous study (13), players in this study weighing more than 100.5 kg (median weight) were estimated to perform their NFL-225 repetitions at 44.6% (621.2%) of 1RM and produce approximately 17.3 (65.9) repetitions, whereas those less than 100.5 kg would perform at 67.9% (615.2%) of 1RM and produce approximately 10.4 (65.9) repetitions. In an earlier study on Division II players (23), the correlation between body weight and repetitions performed at 110% of body weight was nonsignificant (r = 0.17). Furthermore, there was no significant difference between light (,91.0 kg) and heavy (.91.0 kg) players in volume load (load 3 repetitions) estimated using a RepWt of 110% of
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 body weight. A more recent study of Division II players produced similar results (3), noting a nonsignificant correlation between body weight and repetitions at 70% (r = 20.19), 80% (r = 20.11), and 90% of 1RM (r = 20.16). In addition, volume load did not change significantly at each %1RM after a resistance training program that emphasized heavy loads and few repetitions (3). There was also no significant difference between weaker (,113.6 kg) and stronger (.136.4 kg) players in repetitions at each %1RM. However, stronger players achieved significantly greater volume loads at each %1RM than weaker players because of the greater RepWt for each %1RM. This contradicted earlier work on average subjects where the weaker individuals appeared to have an advantage in work performed (1). Although mass has be shown to be significantly related to the NFL-225 test and 1RM (13), statistically controlling for body weight reduced the correlation between 1RM and NFL-225 repetitions (r = 0.96) only slightly (r = 0.94). Therefore, as suggested by Naclerio et al. (18) and from the present observations, the goal of a training program to increase an absolute endurance task, such as the NFL-225 test, should be to maximize a player’s absolute strength level or perhaps his body mass. Because of the use of the NFL-225 test as the exclusive measure of upper-body strength performance at the NFL combine (21,22,24), players preparing to perform well in that event often alter their training to produce greater endurance in the bench press than strength improvement. Indeed, anecdotal evidence from various fitness organizations specializing in preparing players for NFL combine performance suggests that a greater emphasis on high repetitions in the bench press, in conjunction with loaded and unloaded push-ups, should be used. A limited amount of research has focused on high-repetition, low-load training as a means of strength improvement (1,6,10). Given the overwhelming amount of research supporting heavy-load, low-repetition training for strength improvement and the strong connection between absolute strength and repetition performance, it is unlikely that major college strength and conditioning programs should alter their resistance training programs to enhance NFL-225 performance without strong support for the effect of such a change on upper-body strength.
PRACTICAL APPLICATIONS The NFL-225 test is widely used in college football under the supposition of measuring muscular strength in the upper body. Previous studies have observed high correlations between 1RM bench press and the number of repetitions completed in the NFL-225 test while noting an increasing prediction bias at the upper end of the repetition continuum (14–17). This study found that the NFL-225 test is highly reproducible leading to the identification of a smallest worthwhile change of 3 repetitions to denote a significant change in performance greater than between-day variability (1–2 repetitions). Throughout an athlete’s career, it is likely
| www.nsca.com
that increases in absolute or relative strength might be one of the most beneficial factors for improving sports performance (11,12,28). However, considering the importance of showing improvement during a player’s college career and owing to the impact of performance scores at the NFL combine on professional draft status (24), one must consider the degree to which training specifically for this test becomes a priority.
REFERENCES 1. Anderson, T and Kearney, JT. Effect of three resistacne training programs on muscular strength and absolute and relative endurance. Res Q Exerc Sport 53: 1–7, 1982. 2. Bland, JM and Altman, DG. Statistical method of assessing agreement between two methods of clinical measurement. Lancet 327: 307–310, 1986. 3. Brechue, WF and Mayhew, JL. Upper-body work capacity and 1RM prediction are unaltered by increasing muscular strength in college football players. J Strength Cond Res 23: 2477–2486, 2009. 4. Brown, BR, Williams, T, Randolph, KL, and Mayhew, JL. Relationship of anthropometric dimensions to 1-RM bench press and NFL-225 test in college football players. Mo J Hlth Phys Educ Rec Dance 23: 36–45, 2013. 5. Chapman, PP, Whitehead, JR, and Binkert, RH. The 225-lb reps-to-fatigue test as a submaximal estimate of 1RM bench press performance in college football players. J Strength Cond Res 12: 258– 261, 1998. 6. Goto, K, Nagasawa, M, Yanagisawa, O, Kizuka, T, Ishii, N, and Takamatsu, K. Muscular adaptations to combinations of high- and low-intensity resistance exercise. J Strength Cond Res 18: 730–737, 2004. 7. Hetzler, RK, Schroeder, BL, Wages, JJ, Stickley, CD, and Kimura, IF. Anthropometry increases 1 repetition maximum predictive ability of NFL-225 test for division IA college football players. J Strength Cond Res 24: 1429–1439, 2010. 8. Hopkins, WG. Measures of reliability in sports medicine and science. Sports Med 30: 1–15, 2000. 9. Hopkins, WG. Bias in Bland-Altman but not regression validity analyses. Sportscience 8: 42–46, 2004. 10. Jackson, NP, Hickey, MS, and Reiser, RF II. High resistance/low repetition vs. low resistance/high repetition training: Effects on performance of trained cyclists. J Strength Cond Res 21: 289–295, 2007. 11. Kraemer, WJ and Fleck, SJ. Optimizing Strength Training: Designing Nonlinear Periodization Workouts. Champaign, IL: Human Kinetics, 2007. 12. Kurz, T. Science of Sports Training. Island Pond, VT: Stadion Publishing, 1991. 13. Mann, JB, Stoner, JB, and Mayhew, JL. NFL-225 test to predict 1RM bench press in NCAA division I football players. J Strength Cond Res 26: 2623–2631, 2012. 14. Mayhew, JL, Jacques, JA, Ware, JA, Chapman, PP, Bemben, MG, Ward, TE, and Slovak, JP. Anthropometric dimensions do not enhance 1-RM prediction from the NFL-225 test in college football players. J Strength Cond Res 18: 572–578, 2004. 15. Mayhew, JL, Srnka, S, Getty, C, Ball, A, and Jacques, JA. Evaluation of the NFL-225 test for predicting one repetition maximum bench press in small-college football players. Mo J Hlth Phys Educ Rec Dance 20: 67–76, 2010. 16. Mayhew, JL, Ware, JS, Bemben, MG, Wilt, B, Ward, TE, Farris, B, Juraszek, J, and Slovak, JP. The NFL-225 test as a measure of bench press strength in college football players. J Strength Cond Res 13: 130–134, 1999. VOLUME 28 | NUMBER 5 | MAY 2014 |
1431
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
NFL-225 Reliability 17. 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.
23. Schell, J, Ware, JS, and Mayhew, JL. Muscular endurance performance relative to body weight to predict 1-RM bench press in college football players. Iowa Assoc Hlth Phys Educ Rec Dance J 32: 30–31, 1999.
18. Naclerio, FJ, Colado, JC, Rhea, MR, Bunker, D, and Triplett, NT. The influence of strength and power on muscle endurance test performance. J Strength Cond Res 23: 1482–1488, 2009.
24. Sierer, SP, Battaglini, CL, Mihalik, JP, Shields, EW, and Tomasini, NT. The National Football combine: Performance differences between drafted and nondrafted players entering the 2004 and 2005 drafts. J Strength Cond Res 22: 6–12, 2008.
19. Phillips, WT, Batterham, AM, Valenzuela, JE, and Burkett, LN. Reliability of maximal strength testing in older adults. Arch Phys Med Rehabil 85: 329–334, 2004.
25. Stone, MH, Sands, WA, Pierce, KC, Newton, RU, Haff, GG, and Carlock, J. Maximum strength and strength training—A relationship to endurance? Strength Cond J 28: 44–53, 2006.
20. Rhea, MR. Determining the magnitude of treatment effects in strength training research through the use of the effect size. J Strength Cond Res 18: 918–920, 2004.
26. Weir, JP. Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res 19: 231– 240, 2005.
21. Robbins, DW. Positional characteristics of players drafted into the National Football League. J Strength Cond Res 25: 2661–2667, 2011.
27. Whisenant, MJ, Panton, LB, East, WB, and Broeder, CE. Validation of submaximal prediction equations for the 1 repetition maximum bench press on a group of collegiate football players. J Strength Cond Res 17: 221–227, 2003.
22. Robbins, DW. Relationships between national football league combine performance measures. J Strength Cond Res 26: 226–231, 2012.
1432
the
28. Zatsiorsky, VM and Kraemer, WJ. Science and Practice of Strength Training. Champaign, IL: Human Kinetics, 2006.
TM
Journal of Strength and Conditioning Research
Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.
