Secondary Logo

Journal Logo

NFL-225 Test to Predict 1RM Bench Press in NCAA Division I Football Players

Mann, J. Bryan; Stoner, Josh D.; Mayhew, Jerry L.

Journal of Strength and Conditioning Research: October 2012 - Volume 26 - Issue 10 - p 2623–2631
doi: 10.1519/JSC.0b013e31826791ef
Original Research
Free

Mann, JB, Stoner, JD, and Mayhew, JL. NFL-225 test to predict 1RM bench press in NCAA Division I football players. J Strength Cond Res 26(10): 2623–2631, 2012—The National Football League (NFL)-225 test has gained popularity for assessing muscular performance among college football programs. Although the test is a measure of absolute muscular endurance, it was reputed to be highly correlated with maximum muscular strength. The purposes of this study were to assess the predictive potential of the NFL-225 test for estimating 1 repetition maximum (1RM) bench press performance in National Collegiate Athletic Association Division I college football players and to evaluate the accuracy of previous NFL-225 prediction equations. Players (n = 289) in a successful Division I program were assessed over a period of 5 years for 1RM bench press and repetitions completed with 102.3 kg (225 lb). Test sessions occurred within 1 week of each other during the off-season training period. In a validation group (n = 202), repetitions were significantly correlated with 1RM (r = 0.95), producing a prediction equation (1RM [kg] = 103.5 + 3.08 Reps) with a standard error of estimate = 6.4 kg (coefficient of variation = 4.3%). In a randomly selected cross-validation group (n = 87), the new equation nonsignificantly underpredicted by 0.9 ± 7.2 kg produced a high correlation with actual 1RM (intraclass correlation coefficient [ICC] = 0.967), had a limit of agreement of −15.0 to 13.2 kg, and predicted 69% of the group within ±4.5 kg of their actual 1RM. The best previous equation was that of Slovak et al., which was nonsignificantly underpredicted by −0.5 ± 6.7 kg, produced a high correlation with actual 1RM (ICC = 0.975), and predicted 68% of the group within ±4.5 kg of their actual 1RM. The new NFL-225 test seems to be a reasonable predictor of 1RM bench press in Division I players but should be further assessed on players from other high-level programs.

1Athletic Performance Department, University of Missouri, Columbia, Missouri

2Athletic Department, Coastal Carolina University, Conway, South Carolina

3Human Performance Laboratory, Truman State University, Kirksville, Missouri

4Physiology Department, A. T. Still University of Health Sciences, Kirksville, Missouri

Address correspondence to Dr. Jerry L. Mayhew, jmayhew@truman.edu.

Back to Top | Article Outline

Introduction

There is little doubt that American football requires a high degree of muscular strength. For many years, the maximal dynamic strength capability of most players has been assessed by a 1 repetition maximum (1RM). The 1RM may be assessed in a slow controlled movement such as a bench press or squat or in an explosive manner such a push press, power clean, or Olympic snatch. Although there is little doubt that these maneuvers have the ecological validity for evaluating the individual's ability to exert maximal force in one full-range movement, there may be a degree of concern over risk of musculoskeletal injury when the system is subjected to such high loads (11). Although safety records do not indicate a high rate of injury when individuals handle heavy weights (11), nonetheless, a great many individuals have sought an approach that would place less of a maximal stress on body structures during strength evaluation, be accurate, and time efficient.

The accepted measure of strength and muscular endurance in the National Football League (NFL) Combine is the NFL-225 test (14,16,17). This test requires the player to perform as many bench press repetitions as possible using a load of 102.3 kg (225 lb). The origin of the NFL-225 test is undocumented, and the test seems to have no statistical analysis at the professional level. However, because many National Collegiate Athletic Association (NCAA) Division I football teams are primary suppliers for players in the NFL, the NFL-225 test has grown in popularity in recent years among strength and conditioning specialists to evaluate collegiate players. The test has extensive analysis among college players at the Division II (1,6,9) and Division III (7) levels but limited examination at the Division I level (3). Although each of these studies has produced a high correlation between the 1RM predicted from NFL-225 repetitions and the actual 1RM performance (r > 0.93), there exists a larger standard error of estimate (SEE) and degree of heteroscedasticity for the regression equations, meaning the predicting accuracy of repetitions to estimate strength decreases as the number of repetitions increases. Several studies have illustrated this drop in accuracy by a decrease in the correlation between 1RM and repetitions when more than 10 repetitions were completed and by an increase in the SEE (1,6,8,9).

Several studies have evaluated the feasibility of adding anthropometric dimensions to NFL-225 repetitions in an effort to reduce the prediction error in 1RM, especially at the upper end of the repetition continuum. Whisenant et al. (18) noted that the addition of demographic variables such as age, height, weight, %fat (predicted using a 7-site skinfold equation), and fat-free weight (FFW) to NFL-225 repetitions increased the variance accounted for (R2 × 100) among various prediction equations when applied to Division I-AA players. When they used previously developed NFL-225 equations, however, increases in variance accounted for were only 1.1 to 1.7%. Furthermore, they noted that after ethnicity was selected by linear regression, the other variables made no significant contribution to reduction in the variance accounted for. Interestingly, Whisenant et al. (18) used only FFW and height to produce a new NFL-225 equation. In an analysis of Division II players (6), application of the Whisenant et al. (7) equation reduced the variance accounted for by only 4%, indicating little benefit to taking the additional time to estimate body composition. Furthermore, specific variables such as flexed arm muscle cross-sectional area, chest circumference, and arm length did not add to the variance accounted for in NFL-225 equations (6). Hetzler et al. (3) are the only investigators to evaluate the contributions of these demographic variables to 1RM prediction from NFL-225 repetitions in Division I players. They suggested that the arm circumference and arm length improved the variance accounted for by 3%, amounting to an improvement in the SEE of 1.4 kg or 7.7%. The authors did point out that owing to the small sample size and number of variables used, the regression coefficients of their equations may not be stable and widely applicable to other Division I players.

Therefore, it seems reasonable to employ a larger sample size of high-level Division I players to determine the applicability of the NFL-225 concept to predicting 1RM bench press strength. If this technique could produce acceptable results in major college players, it could simplify the strength and conditioning specialists' job of evaluating upper-body performance in players by allowing reliance on one test to both evaluate strength levels and serve as the basis for designing periodized strength programs. Thus, the purposes of this study were to assess the predictive potential of the NFL-225 test for estimating the 1RM bench press performance in NCAA Division I college football players and to evaluate the accuracy of previous NFL-225 equations.

Back to Top | Article Outline

Methods

Experimental Approach to the Study

The use of the NFL-225 test at the Division I level of college football is wide spread despite a limited amount of scientific analysis of its accuracy in predicting maximal strength in players of that caliber. This study sought to evaluate the accuracy of the NFL-225 repetition test for predicting 1RM bench press performance in a large sample of players in a major Division IA football program. Each subject was tested during the off-season conditioning program for the maximum number of repetitions completed using 102.3 kg and for a 1RM bench press, with each test occurring within 1 week of each other. Simple measures of body composition were estimated from body mass index (BMI) to assess the potential of selected demographic variables for reducing the prediction error associated with the NFL-225 test. Multiple regression analysis was used to determine prediction equations and amount of variance accounted for in estimating 1RM strength from the maximum number of repetitions completed. In addition, the accuracy of previously published NFL-225 equations was evaluated for players of this caliber.

Back to Top | Article Outline

Subjects

The subjects (n = 289) of this study were members of a highly successful Division IA program that had a previous 4-year record of 40 wins vs. 14 losses and 4 bowl game appearances. Subjects were divided into 3 player groups depending on the playing position: Big group (n = 116) were offensive line and defensive tackles; Mid group (n = 93) were running backs, tightends, linebackers, defensive ends, and quarterbacks; and Skill group (n = 80) were wide receivers and defensive backs. Subjects were also divided into 3 repetition groups: group 1 = 1 to 10 repetitions (n = 92), group 2 = 11 to 19 repetitions (n = 144), and group 3 = ≥20 repetitions (n = 53). In addition, subjects were divided into 4 weight groups: group 1 = ≤88.6 kg (n = 70), group 2 = 88.7 to 99.1 kg (n = 75), group 3 = 99.2 to 113.6 kg (n = 71), and group 4 = ≥113.7 kg (n = 73). Seventy percent of the subjects (n = 202) were randomly selected as a validation group for construction of a new NFL-225 prediction test to estimate 1RM bench press. The remaining players (n = 87) were used as a cross-validation group to evaluate any newly developed equations. Players had been involved in resistance training programs for 5 to 8 years. Demographic and performance variables for the subjects are presented in Tables 1 and 2. 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.

Table 1

Table 1

Table 2

Table 2

Back to Top | Article Outline

Procedures

1 Repetition maximum Test

The 1RM bench press required the player to lower the bar slowly to touch the chest before being pressed immediately to full arms' extension in the standard “touch-and-go” method. The head, shoulders, and buttocks remained in contact with the bench throughout the lift. Each player was allowed to warm up according to the personal preferences using lightweights of approximately 60 to 80% of estimated 1RM. Testing protocol allowed the player to select a starting weight that would allow completion of one repetition. Depending on the player's perception of difficulty of that lift, weight was added, and after a minimum of 5 minutes rest, a second repetition was attempted. The objective was to have most players reach their 1RM within 3 to 5 attempts. Standard Olympic bars and plates were used for all lifts, and the player used a grip of their preference (typically 15–35 cm greater than shoulder width). Strong verbal encouragement was provided by coaches and players during the lift. All testing was performed between 1500 and 1800 hours. Players were not allowed to perform the test if they had any upper-body injury within several months of the test date. A registered dietitian provided regular advice to ensure adequate nutrition for each player. Players were encouraged to maintain proper hydration and were not allowed to perform resistance training or testing without the possession of their personal water bottle at all times.

Back to Top | Article Outline

National Football League-225 Test

During the week after the 1RM testing and at the same time of the day, each player performed the NFL-225 test using a load of 102.3 kg (225 lb) attempting to complete as many repetitions as possible without pause. After individual warm-ups, the player grasped the bar at the same position used during the 1RM procedure. No mandatory cadence was imposed for the repetition tests, although each player was encouraged to maintain a constant pace of their own choosing. No more than a 2-second pause between each repetition was allowed. The bar was required to touch the chest (but no bouncing off it) on each repetition and be returned to full-arm extension. The head, upper back, and buttocks were required to remain in contact with the bench throughout the test. The test was terminated when the subject could not complete a repetition with proper form. Strong verbal encouragement was provided by coaches and players throughout the test. Reliability for this procedure was high (intraclass correlation coefficient [ICC] = 0.987).

Back to Top | Article Outline

Body Composition Estimation

Body weight and height were used to determine BMI (in kilograms per square meter). Body mass index was then used to estimate %fat accounting to a procedure derived on Division II football players (4). Fat-free weight was estimated using height and weight from an equation developed from the same sample of Division II players (4).

Back to Top | Article Outline

Statistical Analyses

One-way analysis of variance (ANOVA) was used to assess differences among position groups, with Bonferroni post hoc follow-up testing where significance was noted. Pearson correlations were used to determine the relationship among selected variables. Multiple linear regression analysis was used to determine the predictive potential of NFL-225 repetitions and other selected variables to estimate 1RM bench press. Previously reported NFL-225 equations (Table 3) were evaluated for accuracy in estimating 1RM using repeated-measures ANOVA and ICC. Constant error was assessed as the difference between predicted and actual 1RM performances. Total error was calculated as follows:

. The %total error was calculated as follows: Total Error/Actual 1RM Mean × 100. The significance level was set at p ≤ 0.05.

Table 3

Table 3

Back to Top | Article Outline

Results

Significant differences among physical and performance characteristics of the 3 position groups are shown in Table 1. The Big group typically had significantly higher or lower values for most variables compared with the Mid or Skill groups. All demographic variables were significantly correlated with 1RM, accounting for 3 to 34% of the variance in 1RM bench press. Body weight was positively correlated with absolute strength (r = 0.53, p < 0.001) but negatively correlated with relative strength (r = −0.60, p < 0.001), indicating that because players got larger, they tended to be able to lift more total weight but less weight per kilogram of body weight.

Significant differences among the 4 body weight groups are shown in Table 2. The lighter groups tended to be more linear than the heavier groups, with lower BMIs. The lighter groups had lower absolute 1RM values but higher relative values. The NFL-225 load represented a greater %1RM and produced fewer repetitions in the 2 lighter groups than in the 2 heavier groups. Comparisons among the repetition groups indicated that players completing 10 or less repetitions were shorter than those completing ≥20 repetitions but not different from those completing 11 to 19 repetitions. All 3 repetition groups were significantly different from each other in body mass, 1RM, and %1RM represented by the 102.3 kg load. Relative strength (1RM kg−1) for players completing 10 or less repetitions was significantly lower than for players completing 11 to 19 and ≥20 repetitions, with the latter groups differing nonsignificantly.

Table 4 illustrates the cross-validation of previously developed NFL-225 equations. Five of the nine equations produced predicted values that were not significantly different from the actual 1RM. Four equations produced more than 60% of the players with predicted 1RM values within ±4.5 kg (3% error) of their actual 1RM. In the equation producing the greatest number of players within ±4.5 kg of their actual 1RM (6), there was a small positive correlation (r = 0.19, p < 0.01) between body mass and the difference between predicted and actual 1RM values, but this accounted for only 4% of the common variance between the 2 variables. In addition, there was a small positive correlation (r = 0.16, p < 0.01) between the number of repetitions completed and the difference between predicted and actual 1RM values, indicating somewhat less predictive accuracy at the upper end of the repetition continuum (Figure 1).

Table 4

Table 4

Figure 1

Figure 1

Linear regression was used to develop prediction equations to estimate 1RM from NFL-225 repetitions on the validation group (Table 5). The slopes and intercepts for individual position groups were not significantly different and hence could be represented by an overall regression line for the entire sample (Figure 2). The addition of age, height, weight, FFW, or %fat made no significant contribution to the prediction of 1RM. Thus, NFL-225 repetitions seemed to be the major variable contributing to the accurate prediction of 1RM bench press in Division IA football players.

Table 5

Table 5

Figure 2

Figure 2

Cross-validation of the various NFL-225 equations is shown in Table 6. Cross-validation correlations were high for every category and the composite group. Predicted 1RM values within ±4.5 kg of actual values were prevalent for players in the Mid and Skill groups, players completing ≤10 repetitions, and moderate weight players (i.e., 88–99 kg). The percent difference between predicted and actual 1RM values was significantly different between the lightest weight group (group 1 = −3.4 ± 3.4%) and the heaviest weight group (group 4 = 0.8 ± 5.4%), but the middle weight groups (group 2 = 0.2 ± 5.8% and group 3 = −1.3 ± 3.7%) were not significantly different from either extreme. For the entire group, the percent difference was −0.4 ± 4.7% with a %total error of 4.9%. Prediction errors were greater because the number of repetitions completed increased (Figure 3).

Table 6

Table 6

Figure 3

Figure 3

Back to Top | Article Outline

Discussion

This is the first NFL-225 study to used a large sample of top-level college players to assess the accuracy of predicting 1RM bench press from repetitions completed using an absolute load of 102.3 kg (225 lb). The prediction equation produced in this study was similar to others developed on players at various levels of competition, indicating a consistency in the relationship between maximum repetitions with submaximal loads and peak strength. Four of the previous NFL-225 equations developed on less elite players predicted over 60% of the sample within ±4.5 kg, with most of the agreement coming when 10 or fewer repetitions were completed. Cross-validation of the newly developed equation on a random sample of elite players predicted 67% within ±4.5 kg of their actual 1RM. Prediction accuracy tended to be greater in players completing ≤10 repetitions, which agreed with previous studies (1,3,6,9,10,15). This is also the first study to indicate that mid-range body weights (88–99 kg) tended to have their 1RM predicted with greater accuracy than did lighter or heavier players.

Several studies have suggested that including various anthropometric dimensions would reduce the error associated with predicting the 1RM from NFL-225 repetitions alone (3,18). Whisenant et al. (18) used a small number (n = 69) of Division IAA players, whereas Hetzler et al. (3) used Division I players (n = 87). Each study indicated different anthropometric variables contributed to increased amount of variance accounted for, making it difficult to decide exactly which dimensions were better suited to reducing the prediction error. Furthermore, neither study provided a detailed analysis of how much the prediction error was reduced nor if the reductions were uniform across players of all sizes. Using height and weight on the current sample to estimate FFW (4) allowed the determination of 1RM from the equation of Whisenant et al. (18) but produced only 52% of the current sample with estimates within ±4.5 kg of their actual 1RM. This was substantially less when only the NFL-225 repetitions were used. Furthermore, the positive correlation (r = 0.20) between body mass and the difference between the Whisenant et al. (18) predicted 1RM and actual 1RM indicated a heterogeneity in the distribution of the values, with larger individuals having greater error.

The equation of Hetzler et al. (3) was the only published equation on a comparable sample of elite Division I players. The size and strength performances of their players were comparable to those of the current study. However, no more than 35% of their subjects were predicted with an error smaller than ±4.5 kg by any of the previously developed NFL-225 equations. In their small cross-validation sample (n = 31), 52% of their players had predicted 1RM values within ±4.5 kg of their actual 1RM when using only NFL-225 repetitions, which was increased to 65% when their anthropometric equation was used. In the current sample, the basic equation (without anthropometric variables) developed by Hetzler et al. resulted in 90% of the players with predicted 1RM values below actual 1RM values by an average of 8.1 kg (±5.3 kg). Hetzler et al. (3) commented that it seemed surprising that anthropometric dimensions did not contribute to NFL-225 equations in the previous studies. The probable explanation for the lack of major contribution by anthropometric dimensions might be based on the procedure involved in step-wise regression. If there is a high degree of collinearity among variables, the best one will be selected first and variables related to the selected variable will be eliminated from the prediction. Thus, the lack of major contributions by anthropometric variables to reducing the prediction error in both college (6) and high school players (5) when using absolute loads would suggest that little is to be gained by making these additional measurements. Furthermore, the additional time and manpower typically required to measure anthropometric dimensions might defeat one purpose of using the NFL-225 test by requiring more time for measurement without gaining substantial error reduction.

The current study exhibited the same problem as previously noted in all the other studies using an absolute load to estimate 1RM: progressively greater prediction error as the number of repetitions increased beyond 10. In the cross-validation group, the new equation produced 86% of players performing ≤10 repetitions with prediction errors of ≤4.5 kg, which dropped to 61% for players completing 11 to 19 repetitions and 47% for players completing ≥20 repetitions. Among the 4 body weight classifications, the lighter 2 groups produced more of the sample (73%) with ≤4.5 kg error than the heavier 2 groups (59%). Applying various combinations of simple anthropometric variables such as height, weight, and BMI did not alter the prediction accuracy in any of the weight groups.

In conclusion, it would seem that the use of NFL-225 repetitions to estimate 1RM bench strength has statistical support at all levels of college football. The current equation on elite Division I players follows the same trends as the previous research indicating that individuals performing fewer repetitions (≤10) seem to have better prediction accuracy than players performing more than 20 repetitions. Because the objective of the NFL-225 test is not necessarily to estimate maximal strength among professional candidates, there is little likelihood of the test being changed in the foreseeable future to a more equitable predictor for all size players. Previous pilot work has suggested that perhaps a weight equivalent to 110% of a player's body weight would provide a better repetition range for estimating 1RM (13). However, such a suggestion would have to be validated on elite players at the Division I level. The performance on the NFL-225 test, for both drafted and free-agent players in the professional ranks, is one of the important criteria contributing to being selected for and remaining with a professional team (14,16).

Back to Top | Article Outline

Practical Applications

The findings of this study suggest that the NFL-225 test may be an acceptable approach to estimating 1RM bench press strength in Division I college football players. Although the 1RM determination is considered safe in most cases (11), injuries have occurred which could hamper training and playing time for significant players (2,12). Although injuries can occur when attempting maximal repetitions with submaximal loads, emphasis on proper form and maintenance of careful spotting could make the NFL-225 test a quick and safe method of assessing upper-body strength in elite players.

The new equation presented in this study should estimate a large proportion of a Division I team with an error of <5%. Because of the heteroscedasticity that occurs at the upper end of the repetition continuum, the prediction error may increase from 6 to 12% where players perform more than 20 repetitions. Of the players performing more than 20 repetitions, only a small proportion was underpredicted (16%) or overpredicted (16%) by more than 6%. Those players who were underpredicted had a significantly higher bench press per kilogram values than those who were overpredicted. This fact may aid strength and conditioning specialists in making better application of predicted 1RM performance for establishing periodized resistance programs throughout the year.

Back to Top | Article Outline

References

1. Chapman PP, Whitehead JR, 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.
2. Connell DA, Potter HG, Sherman MF, Wickiewicz TL. Injuries of the pectoralis major muscle: Evaluation with MR imaging. Radiology 210: 785–791, 1999.
3. Hetzler RK, Schroeder BL, Wages JJ, Stickley CD, 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.
4. Mayhew JL, Amundson J, Painter K, Landram M, Jorn L. Comparison of %fat and BMI for determining obesity in college football players [abstract]. J Strength Cond Res 14: 433, 1999. Abstract.
5. Mayhew JL, Hill SP, Thompson MD, Johnson EC, Wheller L. Using absolute and relative muscle endurance to estimate maximal strength in young athletes. Int J Sports Physiol Perform 2: 305–314, 2007.
6. Mayhew JL, Jacques JA, Ware JA, Chapman PP, Bemben MG, Ward TE, 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.
7. Mayhew JL, Srnka S, Getty C, Ball A, Jacques JA. Evaluation of the NFL-225 test for predicting one repetition maximum bench press in small-college football players. Missouri J Health Phys Educ Rec Dance 20: 67–76, 2010.
8. Mayhew JL, Ware JS, Bemben MG, Wilt B, Ward TE, Farris B, Juraszek J, 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.
9. Mayhew JL, Ware JS, Cannon K, Corbett S, Chapman PP, Bemben MG, Ward TE, Farris B, Juraszek J, 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.
10. Mayhew JL, Ware JS, Clemens CT, Coco B. Accuracy of prediction equations for determining 1-RM bench press, squat, and deadlift in college football players. J Human Mov Studies 40: 423–436, 2001.
11. Mazur LJ, Yetman RJ, Risser WL. Weight-training injuries: Common injuries and preventative methods. Sports Med 16: 57–63, 1993.
12. Petilon J, Carr DR, Sekiya JK, Unger DV. Pectoralis major muscle injuries: Evaluation and management. J Am Acad Orthop Surg 13: 59–68, 2005.
13. Schell J, Ware JS, Mayhew JL. Muscular endurance performance relative to body weight to predict 1-RM bench press in college football players. Iowa Assoc Health Phys Educ Rec Dance J 32: 30–31, 1999.
14. Sierer SP, Battaglini CL, Mihalik JP, Shields EW, 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.
15. Slovak JP, Ward TE, Mayhew JL. Cross-validation of NFL-225 tests for predicting 1 repetition maximum bench press performance for a college football team [abstract]. J Strength Cond Res 14: 433, 1999. Abstract.
16. Robbins DW. The National Football League (NFL) combine: Does normalized data better predict performance in the NFL draft? J Strength Cond Res 24: 2888–2899, 2010.
17. Robbins DW. Relationships between National Football League combine performance measures. J Strength Cond Res 26: 226–231, 2012.
18. Whisenant MJ, Panton LB, East WB, 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.
Keywords:

muscular strength; absolute muscular endurance; strength; prediction

Copyright © 2012 by the National Strength & Conditioning Association.