The Validity and Reliability of the 1RM Bench Press Using Chain-Loaded Resistance : The Journal of Strength & Conditioning Research

Secondary Logo

Journal Logo

Original Research

The Validity and Reliability of the 1RM Bench Press Using Chain-Loaded Resistance

McCurdy, Kevin; Langford, George; Jenkerson, David; Doscher, Michael

Author Information
Journal of Strength and Conditioning Research: May 2008 - Volume 22 - Issue 3 - p 678-683
doi: 10.1519/JSC.0b013e31816a6ce0
  • Free



Using chain-loaded barbells for resistance has recently become popular as a method to potentially enhance strength and power improvement. Chains are loaded on a free-weight bar and combined with traditional plates or added to the bar as the entire load. Anecdotal evidence suggests that this type of training can increase strength and power and reduce joint stress by providing an accommodating resistance during exercises such as the bench press (15). Research is needed to investigate these reported benefits. Significant reduction of joint stress and improved rate of force development with higher bar velocities associated with lower load through a portion of the range of motion are potential benefits of training with chains providing the entire resistance. Determining valid and reliable measures of maximal strength using chain-loaded resistance is also needed to assign the appropriate resistance for the prescribed intensity and to monitor strength improvement with meaningful results to determine adjustments to the training loads as strength improves.

Several studies have found reliable and valid measures of maximal strength using different types of isokinetic and isotonic bench press machines (1,4,7,9,12,14,18,19). These machines balance and stabilize the resistance while isokinetic devices control the velocity, which likely enhances the reliability. Conversely, chains provide a variable resistance throughout the range of motion. The chains remain hanging from metal hooks that are secured to the bar during the lift while a portion of the load is in constant contact with the floor (Figure 1). During the descent of the bench press, the resistance is unloaded as more links in the chain rest on the floor; the load increases during the ascent. The movement of the chains presents potential instability that requires an effort to control. This instability may reduce the reliability during strength assessment as the precision of the forces applied is essential to achieve a reliable measure of maximal strength.

Figure 1:
Chain-loaded bench press.

According to the force-velocity relationship as the required force increases, the velocity of movement decreases (5); therefore, during the ascent of the chain-loaded, free-weight bench press (CBP), the speed of the movement decreases. The unfamiliarity with the variable load and subsequent change in velocity may affect the reliability of the maximal strength test. In addition, with the implementation of unfamiliar exercises, a certain amount of familiarization is required to minimize systematic error due to a learning effect during and after baseline testing (17); otherwise, determined improvements during posttesting may be inflated. Weiss et al. (19) found high reliability (0.95-0.99) for force and power obtained from an Ariel dynamometer bench press across multiple controlled velocities. Similar results were found with this equipment and test protocol during the squat (20). In contrast, Blazevich and Gill (4) found high reliability for peak force with an unfamiliar semi-prone squat machine at 70% of isometric force for a bilateral and unilateral test (r = 0.92 and r = 0.91, respectively), but reliability was somewhat reduced for the bilateral test (r = 0.85) and significantly reduced for the unilateral test (r = 0.57) at 40%. Limited available time to make alterations in the movement sequence based on feedback from proprioceptors as velocity of the movement increased was reported as a possible explanation for the reduced reliability. The effect of a variable, free-weight-loaded barbell with subsequent changes in velocity on the reliability of a maximal bench press strength test is currently unknown.

The most common accepted and utilized measurement of upper-body maximal strength for athletic populations is the plate-loaded, free-weight bench press (PBP). However, with the inclusion of new and innovative methods of free-weight training, often to compliment limitations of dynamic constant external resistance such as traditional plate-loaded exercises (6), appropriate methods of assessment must be developed for these non-traditional types of free-weight exercises and tested for validity and reliability. When chains are used during training, the use of chains during assessment is the most appropriate method to monitor improved strength and to achieve specificity between testing and training. Different upper-body, free-weight exercises must be analyzed for validity and reliability using men and women athletes as subjects so that generalizations can be drawn specifically for the populations who train with these exercises. Correlations between the CBP and the PBP would provide insight into the validity of the CBP and provide strength and conditioning specialists with the ability to estimate achievement on the CBP using the free-weight scores when it is desired to limit the time of testing. Furthermore, when performing an unfamiliar strength test, a learning effect could take place. Therefore, the purposes of this study were to determine: 1) the test-retest reliability of the CBP in men and women college basketball players; 2) the relationship between the 1RM strength scores between the CBP and the PBP; and 3) possible differences between the CBP test and retest measures.


Experimental Approach to the Problem

Athletes were chosen as subjects for this study because of increased use of chain-loaded exercises to improve strength and athletic performance. With the use of chains as the entire load added to the bar during training, the most appropriate method of assessment to monitor strength improvement is chain resistance. Data are needed to determine whether strength assessment using chains is reliable and valid. Although not yet determined through research, basketball players may benefit from CBP training by minimizing muscle soreness and shoulder stress with the unloading of resistance during the eccentric phase. As the bar is lowered to a position near the bottom of the lift and closest to the chest, the involved joints are at a mechanical disadvantage. In this position, the anterior tissues of the glenohumeral joint are stretched near the end of the range of motion, increasing stress on this joint (13). Basketball players have longer than average arms and must increase the range of motion of the humerus during horizontal extension to touch the bar to the chest, potentially further increasing the stretch and stress on the shoulder joint. Utilizing only chains to the weight of the bar will significantly reduce the load at bottom of the bench press; therefore, CBP assessment and training with the entire load as chains may be particularly warranted to reduce joint stress while improving strength and power for this group of athletes.

The PBP test was conducted first to determine the 1RM for estimation of the starting weights during CBP testing. Achieving a maximal lift during the PBP demonstrates that this load can be lifted through the position of the least mechanical advantage. An equal weight measured at the top of the CBP will be significantly less in this position of mechanical disadvantage; therefore, we predicted that the CBP 1RM would be significantly higher than the PBP 1RM because of lighter loads encountered through the sticking point, which allowed relatively high velocities to be produced before the heaviest load occurred at the end of the range of motion. For this reason, 105% of each subject's 1RM on the PBP was used for the initial 1RM trial on the CBP test. This was predicted to be a conservative weight that would produce a successful trial. The relationship between the PBP and CBP maximal strength scores was analyzed to determine the validity of the CBP. The CBP was retested after 4 days to determine the reliability of this measure of 1RM strength and possible mean differences between tests.


Nine men and seven women participating in Division II basketball volunteered for this investigation. The mean age, height, and weight were 20.58 ± 1.31 years, 188.24 ± 9.29 cm, 92.07 ± 16.94 kg and 20.42 ± 0.98 years, 175.61 ± 9.32 cm, 73.61 ± 10.80 kg for men and women, respectively. All of the subjects signed written informed consent forms that were reviewed by the institutional review board at the university to ensure the subjects were knowledgeable of the normal risks and procedures involved in the study. This study was conducted after the competitive season with an in-season resistance training program that included PBP training. All of the subjects had 6 months to several years of resistance training experience that consisted primarily of free-weight training with plate-loaded resistance but did not have training experience with chain-loaded resistance.

1RM Test Procedures

Two sessions were devoted to practicing the CBP and PBP technique with light to moderate loads that approached their estimated 1RM and measuring the distance from the bar held in the top position of the bench press to the floor. Seven-foot chains varying in size from 1-in to ¼-in links (Figure 1) were used to add resistance. Each chain was labeled with a different color to identify the correct size chain and resistance. The length of the chains lifted off of the floor for each subject's top position was weighed to determine the load added to the bar. By varying the combinations of the different chains, resistance could be added in approximately 5-lb increments. The chains hung from collared hooks that were secured to the bar. To improve stabilization, a portion of the chains remained in contact with the floor.

The subjects were instructed to take a grip width slightly wider than the shoulders and to lightly touch the chest before returning to the top position while keeping the feet on the floor and the hips, upper back, and head on the bench. All subjects completed the PBP 1RM test followed by the test and retest on the CBP with a minimum of 4 days' rest between each test. A 5-minute warm-up and upper-body stretches were monitored before each test.

All test procedures were monitored by the investigators and provided in clear view for the subjects to read. Before the CBP test, trial loads were calculated in 5% increments from an initial trial of 105% to a possible final trial of 150% of each subject's PBP 1RM. The CBP 1RM was estimated to occur 105-150% of the subject's PBP 1RM. During the CBP tests, the number and color of the chain were called out to load the required weight without the need for calculations during the procedure. After two warm-up sets of five repetitions using light loads, a conservative weight was added to complete two to three repetitions. One hundred five percent of the subjects' PBP 1RM was used as the initial trial on the CBP, which was estimated to be a conservative initial load. Two to three minutes of rest was allowed on each successive 1RM trial with the addition of 5-10% weight depending on the ease of the successful lift. If the subject was unsuccessful, 2.5-5% was subtracted for a final trial (2). A retest on the CBP took place 4 days after the initial CBP test.

Statistical Analyses

SPSS software version 12.0 was used for the statistical analyses. Few studies have analyzed the relative and absolute reliability within a study to determine the reliability of maximal strength scores (7,20). Absolute reliability describes the within-subject variability among trials (3), which was determined by the coefficient of variation (CV). A CV percentage was determined using the formula CV = 100(es − 1), where s is the random error. Random error was calculated from the standard deviation of the trial to trial differences divided by the square root of 2 (8). Intraclass correlation coefficients (ICC) were determined to analyze the test-retest relative reliability to describe the degree of association between the repeated measures of the CBP scores (16). Relative reliability can be high when there is large between-subject variability regardless of the agreement between test-retest scores (12). Consequently, relative and absolute reliability measures were analyzed to provide a more complete assessment of the reliability. Pearson's correlation coefficients were calculated to determine the relationship between the PBP and CBP. Paired t-tests were utilized to evaluate shifts in the mean between the test and retest on the CBP as a result of the potential for a learning effect. The level of significance was set at p ≤ 0.05 for all analyses.


The men's and women's mean 1RM scores on the PBP and CBP are reported in Table 1. High relative reliability (ICC) was found between the test and retest on the CBP for the men (r = 0.99), women (r = 0.93), and combined group (r = 0.99). Absolute reliability was also high, which was concluded from the low CV for the men (14%), women (3.5%), and combined group (2.5%). Significant correlations were found between the PBP and CBP scores for the men (Table 2) and women (Table 3). Similar correlations between the PBP and CBP (r = 0.99) were found for the men and women combined. The men's average 1RM on the CBP from both tests was 26.5% higher than their 1RM PBP strength, whereas the women's 1RM on the CBP was 20.7% higher than their PBP strength. The paired t-test comparing the difference between the test and retest on the CBP revealed a significantly higher (2.57 kg) 1RM on the retest for the men (p = 0.04), but no difference (0 kg) occurred between the tests for the women (p = 1.0).

Table 1:
1RM for the traditional plate-loaded bench press test and chain-loaded bench press test and retest (men n = 9; women, n = 7).
Table 2:
Correlations between the men's 1RM on the chain-loaded bench press and plate-loaded bench press.
Table 3:
Correlations between the women's 1RM on the chain-loaded bench press and plate-loaded bench press.


The results of this study revealed that the 1RM on the CBP can be determined for young men and women athletes with high test-retest reliability. For the men and women, a high degree of association occurred between the repeated measures with a low within-subject variability. Although the subjects in this study were unfamiliar to this variable resistance, the reliability of the 1RM CBP test was very high for the men and women, which was indicated by the high ICC and low percent CV. In a previous study, Blazevich and Gill (4) found significantly reduced reliability in an unfamiliar squat strength test and concluded that unfamiliar tests may reduce the reliability, particularly if the resistance is difficult to stabilize. A unilateral versus bilateral squat in a semi-prone position on a sled apparatus was included to change the level of stability. The loading and unloading of the chains from the bar to the floor seems to create an unstable resistance; however, the movement of the chains did not affect the reproducibility of the strength scores. Two collars containing two hooks on each collar (Figure 1) were needed on each side of the bar to hold the chains and ensure that the chains did not become entangled on the floor, which could increase the instability.

With the lack of previous CBP training experience before the study, the subjects encountered the unfamiliar change in velocity that occurs with the variable load. The bench press is an ascending strength curve, which introduces the possibility of lifting more weight at the end of the repetition if only the last part of the exercise is performed (6). The variable resistance provided with chains attempt to match this strength curve by providing the most resistance at the end of the repetition with concomitant deceleration of the bar. The subjects in this study were athletes with previous PBP training and were familiar with the initial PBP bar acceleration followed by an oscillation and deceleration phase described by Lander et al. (11). Although velocities of the CBP and PBP were not compared in this study, the CBP velocities were likely higher early in the concentric phase with constant deceleration through the range of motion and slowest near the top position. The data signify that the subjects were able to adjust to these differences in velocity dictated by the variable resistance.

The high correlations between the 1RM on the PBP and CBP indicate that the CBP is a valid test of 1RM strength. Although the correlations found for the women were high, they were lower than the correlations determined for the men. The lower correlations may have been a result of differences in confidence with an unfamiliar test and variable load that also varies in stability as the weight is constantly being loaded and unloaded off the floor.

Although not a clinically large difference, the men's mean CBP 1RM significantly increased during the retest, whereas no change took place for the women's CBP 1RM. With prior knowledge of their CBP 1RM, the subjects were able to reduce the number of trials needed to obtain their 1RM by allowing for larger increments between trials within the guidelines of the procedures. The men completed the PBP test with a mean of 4.67 trials and 6.11 and 4.56 trials, respectively, on the CBP test and retest. The women completed the PBP test in 4.43 trials, and a similar number of trials were needed to obtain their 1RM on the CBP test (5.14) and retest (4.43). Fatigue in the first CBP test may have been a factor for the increase in mean strength scores for the men, but contradictory to this explanation, the women's strength did not change with the reduced number of trials in the retest. Increased confidence and motivational differences occurring among tests between the men and women may, in part, explain the differences found. Phillips et al. (12) determined that three familiarization and two test sessions were needed to minimize a learning effect when older men and women were tested for 1RM strength on a plate-loaded bench press machine. However, our data indicate that the two familiarization sessions provided in this study were sufficient for the athletes to control for a learning effect on the CBP. In comparison, Jidovtseff et al. (10) found that one familiarization session that included bench press practice and a 1RM pretest on a Smith machine was sufficient to eliminate a learning effect in subjects who had previous training experience. The CBP provides less stabilization than the Smith machine; thus, two familiarization sessions with one session including loads approaching the estimated 1RM are likely needed to eliminate a learning effect.

Comparing the 1RM differences between the CBP (bar position at the top position) and PBP, the men's and women's CBP 1RM were 27% and 21% higher, respectively. Possible differences in confidence and motivation between the men and women may affect the difference in percentage between the PBP and the CBP. Although the number of years training was similar between the men and women, a difference in previous training experience with various types of free-weight equipment may have produced a difference in confidence between the men and women. The strength differences can be utilized to estimate appropriate initial loads during CBP testing and training loads with prior knowledge of the PBP 1RM. Further comparison of the 1RM between the CBP and PBP with different athletic populations of men and women is warranted.

Practical Applications

Based on the results of this study, as strength and conditioning specialists implement chain-loaded exercises such as the CBP, 1RM strength testing could be administered to monitor training with confidence in obtaining valid and reliable data. Pretests could also take place to determine appropriate loads for specified intensities used during training. Our data indicate that two CBP familiarization sessions are sufficient for men and women athletes who have previous resistance training experience but do not have experience with this type of training to become proficient for testing 1RM strength. The statistically significant shift in the mean of 2.6 kg was clinically small and likely minimized with prior knowledge of the PBP 1RM. These data can be used to determine an appropriate initial weight on the first and subsequent trials during testing.

To administer the CBP test, preplanning work is required to measure each subject's bar distance from the top position to the floor and weigh each chain based on these lengths. Before the test, increments within approximately 5 lb that are near the subjects' estimated 1RM can be calculated by adjusting the combinations of the chains that vary in thickness. Although initially time consuming, these calculations matched with the associated correct combination of chains can be recorded and used for future tests. Labeling the different chains reduces the time involved in loading the bar and enhances identification of the correct chain and load. This preplanning enhances the ease of administering the test and reduces the testing time, which was similar between the CBP and PBP.


The authors would like to extend their gratitude to Mr. Bill Allerheiligen for the chain-loaded 1RM tables and Drs. John Walker and Gail Ryser for their statistical review.


1. Alemany, J, Pandorf, C, Montain, S, Castellani, J, Tuckow, A, and Nindle, B. Reliability assessment of ballistic jump squats and bench throws. J Strength Cond Res 19: 33-38, 2005.
2. Baechle, RE and Earle, RW. Essentials of Strength Training and Conditioning (2nd ed.). Champaign, IL: Human Kinetics; 2000.
3. Birmingham, T, Kramer, J, Speechley, M, Chesworth, B, and MacDermid, J. Measurement variability and sincerity of effort: clinical utility of isokinetic strength coefficient of variation scores. Ergonomics 41: 853-863, 1998.
4. Blazevich, A and Gill, N. Reliability of unfamiliar, multi-joint, uni- and bilateral strength tests: effects of load and laterality. J Strength Cond Res 20: 226-230; 2006.
5. Caiozzo, V, Perrine, J, and Edgerton, V. Training-induced alterations of the in vivo force-velocity relationship of human muscle. J Appl Physiol 51: 750-754, 1981.
6. Fleck, S and Kraemer, W. Designing Resistance Training Programs (3rd ed.). Champaign, IL. Human Kinetics; 2004.
7. Grooten, W, Puttemans, V, and Larsson, R. Reliability of isokinetic supine bench press in healthy women using the ariel computerized exercise system. Scand J Med Sci Sports 12: 218-222, 2002.
8. Hopkins, W. Measures of reliability in sports medicine and science. Sports Med 1: 1-15, 2000.
9. Hortobagyi, T and Katch, F. Reliability of muscle mechanical characteristics for isokinetic and isotonic squat and bench press exercise using a multifunction computerized dynamometer. Res Q Exerc Sport 61: 191-195, 1990.
10. Jidovtseff, B, Croisier, J, Lhermerout, C, Serre, L, Sac, D, and Crielaard, J. The concept of iso-inertial assessment: reproducibility analysis and descriptive data. Isokinetics Exerc. Sci 14: 53-62, 2006.
11. Lander, J, Bates, B, Sawhill, J, and Hamill, J. A comparison between free-weight and isokinetic bench pressing. Med Sci Sports Exerc 17: 344-353, 1985.
12. Phillips, W, Batterham, A, Valenzuela, J, and Burkett, L. Reliability of maximal strength testing in older adults. Arch Phys Med Rehabil 85: 329-334, 2004.
13. Rosenthaul, M. Shoulder savers: alterations of traditional exercises. Strength Cond J 19: 7-10, 1997.
14. Siegel, J, Gilders, R, Staron, R, and Hagerman, F. Human muscle power output during upper- and lower-body exercises. J Strength Cond Res 16: 173-178, 2002.
15. Simmons, L. Bands and chains. Powerlifting USA 22: 26-27, 1999.
16. Stratford, P. Reliability: consistency or differentiating among subjects. Phys Ther 69: 299-300, 1989.
17. Thomas, J and Nelson, J. Research Methods in Physical Activity. Champaign, IL: Human Kinetics; 1996.
18. Verdera, F, Champavier, L, Schmidt, C, Bermon, S, and Marconnet, P. Reliability and validity of a new device to measure isometric strength in polyarticular exercises. J Sports Med Phys Fitness 39: 113-119, 1999.
19. Weiss, L, Fry, A, Gossic, E, Webber, J, and Barrow, E. Reliability of bench press velocity-spectrum testing. Meas Phys Ed Exerc Sci 2: 243-252, 1998.
20. Weiss, L, Relyen, G, Ashley, C, and Propst, R. Reliability of selected measures of musculoskeletal function obtained during closed kinetic chain exercises at multiple velocities. J Strength Cond Res 10: 45-50, 1996.

free-weight; chains; variable resistance

© 2008 National Strength and Conditioning Association