The impact of test loads on the accuracy of 1RM prediction using the load-velocity relationship

The impact of test loads on the accuracy of 1RM prediction using the load-velocity relationship Background: Numerous methods have been proposed that use submaximal loads to predict one repetition maximum (1RM). One common method applies standard linear regression equations to load and average vertical lifting velocity (V ) data developed during squat jumps or three bench press throw (BP-T). The main aim of this mean project was to determine which combination of three submaximal loads during BP-T result in the most accurate prediction of 1RM Smith Machine bench press strength in healthy individuals. Methods: In this study combinations of three BP-T loads were used to predict 1RM Smith Machine bench press strength. Additionally, we examined whether regression models developed using peak vertical bar velocity (V ), peak rather than V , provide the most accurate prediction of Smith Machine bench press 1RM. 1RM Smith Machine mean bench press strength was measured directly in 12 healthy regular weight trainers (body mass = 80.8 ± 5.7 kg). Two to three days later a linear position transducer attached to the collars on a Smith Machine was used to record V and V during BP-T between 30 and 70% of 1RM (10% increments). mean peak Results: Repeated measures analysis of variance testing showed that the mean values for slope and ordinate intercept for the regression models at each of the load ranges differed significantly depending on whether V or mean V were used in the prediction models (P < 0.001). Conversely, the abscissa intercept did not differ significantly peak between either measure of vertical bar velocity at each load range. The key finding in this study was that 1RM Smith Machine bench press strength can be determined with high relative accuracy by examining V and V mean peak during BP-T over three loads, with the most precise models using V during loads representing 30, 40 and 50% peak of 1RM (R = 0.96, SSE = 4.2 kg). Conclusions: These preliminary findings indicate that exercise programmers working with normal healthy populations can accurately predict Smith Machine 1RM bench press strength using relatively light load Smith Machine BP-T testing, avoiding the need to expose their clients to potentially injurious loads. Keywords: Strength assessment, Dynamic strength, Predictive models, Bench press throws Background issues as it is not only time consuming, but the outcome is The quantification of the maximum load that can be lifted effected by factors such as athlete experience, technique, through a full range of motion, or one repetition maximum fatigue and motivation [2]. Traditional 1RM testing is con- (1RM), is fundamental to the design of resistance training sidered to be safe when it is conducted in appropriate set- programs [1]. Typically, 1RM is either measured directly or tings under the supervision of qualified practitioners [3, 4]. calculated indirectly using predictive models. The direct Regardless, this 1RM exposes athletes to large musculoskel- determination of 1RM suffers from a number of pragmatic etal forces, and there is some evidence that 1RM testing can be potentially injurious [5, 6]and mayalsobeimpracti- cal with novices and/or in clinical settings [7]. * Correspondence: msayers@usc.edu.au Indirect methods for 1RM quantification tend to School of Health and Sport Sciences, University of the Sunshine Coast, follow two different protocols, both of which rely on the Maroochydore DC, QLD 4558, Australia Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 2 of 8 use of linear regression modelling. The most common technology that, due to large reductions in pricing, is indirect protocols involve lifting submaximal loads to becoming increasing accessible to strength coaches. Re- failure [7–9], a procedure that is relatively common in searchers report high correlations using this methods trained athletes [10] and in agreement with the ACSM between Load0 and 1RM bench press (r = 0.98, n = 112, guidelines of 8–12 repetition that is often used in a clin- SEE = 4 kg [7%]), although the strength of this relation- ical setting, but rare in everyday activities. Although ship is no doubt influenced by the large range in relative relatively easy to administer, the accuracy of these ‘lift to loads assessed (30 to 95% of 1RM) [16, 22]. Additionally, failure’ models is also influenced by elements such as participants in the study by Jidovtseff and co-workers age, training experience, motivation and lifting tempo [16] were required to always hold the bar (i.e. prevented [7–9, 11]. The prediction of 1RM using these methods from performing a BP-T), which will have a marked ef- appears to be more accurate when heavier loads are used fect on V due to the deceleration of the bar near the mean [12–14], with the optimal number of repetitions for end of the lift [23]. Nevertheless, some questions remain these prediction models being less than 10 [8]. Accord- as to whether V or V provides superior predict- mean peak ingly, these protocols potentially suffer from the same ive measures. Recently, Gracia-Ramos and co-workers limitations associated with 1RM testing, with the need [24] report that V during Smith Machine bench mean to lift high relative loads whilst fatigued. Additionally, press is a superior predictor of 1RM when compared to these lift to failure protocols are also likely to generate V . However, these findings appear to be specific to peak post exercise muscle soreness in novices [15], potentially the testing protocols as these researchers highlight in dissuading them from future exercise participation. another study that V during bench press throws is peak Alternative indirect methods rely on the load-velocity the superior predictor of 1RM [25]. Regardless, V mean [6, 16] or force-velocity [10, 11, 17–19] relationships and and V appear to be greater predictors of optimal load peak linear or quasi linear models to predict 1RM from a for power training than traditional methods that advo- series of maximal effort lifts with submaximal loads. cate percentages of 1RM [26]. These protocols use either isoinertial sensors or linear Nevertheless, the question remains as to the efficacy of position transducers that are attached to the collars or the procedures proposed by Jidovtseff and co-workers bar of training devices like Smith Machines to record [16], particularly when testing novice or inexperienced force, average and/or peak vertical lifting velocity data weight trainers for which higher lifting loads may be from the concentric phase of movements like jump contraindicated. Therefore, it is important to determine squats or bench press throws (BP-T). Although these whether such high relative loads are required during movements are more common in high performance these submaximal test protocols (i.e. up to 95% of 1RM) training programs, the use of a Smith Machine and ap- and which combination of relative loads result in the propriately trained “Spotters” means that these exercises most accurate predictive model of 1RM bench press can be completed safely with novice participants (NB: strength. Accordingly, the purpose of this study was to some Smith Machines contain a pneumatic brake which use the prediction model developed by Jidovtseff and prevents the bar from descending rapidly – hence im- coworkers [16] to determine which combination of three proving exercise safety). Although 1RM data recorded submaximal loads during BP-T result in the most accur- on Smith Machines are typically 10% higher than those ate prediction of 1RM Smith Machine bench press recorded using free weights, there are no significant strength in healthy individuals. We also examined differences between predicted 1RM values when using whether the ability to release the bar during the BP-T these devices [20]. These protocols also have the advan- changes the nature of the prediction model. Addition- tage of being relatively quick to perform as they involve ally, we examined whether V or V provides a mean peak loads between 30 and 80% of 1RM [11, 16, 21] being better prediction of Smith Machine bench press 1RM lifted as rapidly as possible for only 2–4 repetitions. strength in these participants. Accordingly, the overall loading in these protocols is less than ‘lift to failure’ protocols and so the risk of injury Methods may be decreased, particularly when applied to relatively Experimental approach to the problem untrained populations [16]. To determine which combination of three loads during Arguably, the simplest of the load-velocity models [16] BP-T results in the most accurate prediction of 1RM applies standard linear regression equations to load and bench press strength we tested 12 healthy, regular mean vertical propulsive lifting velocity (V ) data weight trainers on two separate occasions. On the first mean from three different loads to develop slope, abscissa occasion 1RM bench press strength was recorded using (Load0) and ordinate (V 0) intercept data. Import- standard procedures and recorded to the nearest 1 kg mean antly, variables such as V or peak vertical velocity [27]. During the second data collection (2–3 days after mean (V ) can be measured using relatively inexpensive the first testing session) participants performed three peak Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 3 of 8 repetitions of BP-T at loads representing 30, 40, 50, 60 [2]. The bar handle was attached permanently to this and 70% of their 1RM. We then processed these braking system, resulting in a total weight of 23 kg. To load-velocity BP-T data using the techniques proposed record the vertical position of the bar a linear position by Jidovtseff and coworkers [16] to determine which transducer (LPT) (WS17KT, ASM, Moosinning, Germany) three load range (30–50% of 1RM, 40–60% of 1RM, was installed on the Smith Machine’s pneumatic brake, 50–70% of 1RM) resulted in the most accurate pre- with data subsequently sampled at 1000 Hz, A/D con- diction of 1RM bench press strength. We also exam- verted and stored on a computer, Subsequent data analysis ined whether BP-T V or V provides a more of the LPT measurement were performed in MATLAB, mean peak accurate prediction of bench press 1RM by comparing with velocity data developed from the raw LPT outputs each of the models developed using these variables. using the first central difference method. The BP-T testing was conducted in accordance with Participants well-established protocols [23, 28, 29] at loads represent- Participants for this study (n = 12) were all recreational ing 30, 40, 50, 60 and 70% relative to 1RM. The execu- weight trainers who had been weight training at least tion order was determined randomly using Microsoft twice a week for a minimum of 1 year (body mass (BM) = Excel in order to avoid possible order effects during the 80.8 ± 5.7 kg, 1RM 84 ± 18 kg, relative 1RM =1.04 BM [i.e. testing session. In order to minimise the effects of relative load is represented as a function of BM]). None of fatigue there were 2–4 min between repetitions with the participants were involved in heavy load strength three repetitions completed at each load. The eccentric training. Participants were informed of the experimental phase was at a self-chosen speed, with the participants procedures and risks and provided their written informed required to wait for the start signal before commencing consent prior to attending several familiarisation sessions. the concentric motion [30]. There was approximately at This research was approved by the institutional Human 2 s pause between the eccentric and concentric phases. Research Ethics Committee (No. 2012-N-10). The V , and V and maximum bar acceleration mean peak were calculated from the first and second differentials of Procedures the linear transducer data. These data were then used to All bench press and BP-T data were collected on a develop a linear regression model for the prediction of standard Smith Machine. This machine was modified 1RM [16] (Fig. 1). We subsequently developed slope, with a custom made magnetic braking system as a safety Load0 and V 0 data for each of these regressions mean mechanism. Once the bar was released this safety mech- models over each of the load ranges (i.e. 30–50% 1RM, anism prevented it from falling back on the participant 40–60% 1RM and 50–70% 1RM). Fig. 1 Sample data from one subject, three loads (solid circles) processed using standard load-velocity techniques [16]. Graph includes the regression line and the calculated peak mean vertical velocity (V 0), theoretical load at 0 m/s (Load0) and average vertical lifting velocity mean (V )at1RM mean Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 4 of 8 Statistical analyses Results The influence of load on the various bar kinematic vari- ANOVA testing indicated that V and V both mean peak ables were determined via a series of repeated measures showed large, significant reductions (P < 0.001, ES > 1.2) analysis of variance (ANOVA) tests. Post-hoc analyses for each respective increase in relative BP-T load, except were undertaken using paired t-Test with Bonferroni for V between 40 and 50% of 1RM (Fig. 2). Results mean corrections. Shapiro-Wilk and Mauchly’stest ofsphericity also showed that the mean values for slope and V 0 mean were applied during all ANOVA testing. Where data at each of the load ranges differed significantly depend- violated the sphericity assumption Greenhouse-Geisser ing on whether V or V were used in the predic- mean peak corrections were applied. The relative magnitude of tion models (Table 1). Conversely, Load0 data did not differences were quantified using standard Cohen’s Effect differ significantly between either bar velocity measures Size (ES) analyses, with the following descriptors used to at each load range. The CV% values range from 7.2 up define the relative magnitude of the ES: < 0.2 = trivial, to 27.5% (Table 2), with the ICC (Table 3) data show 0.2–0.6 = small,0.6–1.2 =medium/moderate,1.2–2.0 = excellent reliability for V at the lightest range weight peak large,and>2.0= very large [31]. The predictive accuracy whereas only moderate reliability was observed for the of the model developed by Jidovtseff and coworkers [16] weights between 40 and 60%. All other cases showed was assessed using the three lightest loads (30–50% 1RM), good reliability. Typically, greater levels of acceptable re- the three middle loads (40–60% 1RM) and the three liability [32] were recorded for V compared to V . peak mean heaviest loads (50–70% 1RM), with these data then com- There were no noticeable differences in any of the pared with the measured 1RM values. Bland-Altman plots models that used V to predict 1RM Smith Machine mean were used to assess whether there were any levels of bias bench press strength (R between 0.85–0.89). Similarly, in any of the models, with simple t-tests used to assess for there were no significant differences between predicted differences between the actual and predicted values. The and actual 1RM Smith Machine (P = 0.21 to 0.95) when coefficient of variance (CV%) and the intra class correla- using V , although the corresponding Bland-Altman mean tions (ICC, 3,1) for the predicted versus the measured plots highlighting some issues with the accuracy of these 1RM were also calculated. Statistical analysis were per- data (Fig. 3). Conversely, there were significant differ- formed using the statistics package SPSS for Windows ences between the predicted and actual 1RM Smith (version 20), with a confidence level of 95%. All data are Machine bench press values when using V at the peak presented at means ±1 standard deviation (SD) unless lightest of the load ranges to (P < 0.001). However, the stated otherwise. predicted 1RM values for V for these light relative peak Fig. 2 Mean (1SD) mean (V ) and peak bar (V ) vertical velocities at each of the relative loads. * Indicates data significantly different mean peak (P < 0.01) than the other loads Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 5 of 8 Table 1 Mean (±1SD) values of the slope, abscissa (Load0) and ordinate (V 0) intercept data for each regression line developed mean using both V and V across the three loading ranges mean peak Variable Percent of 1RM 30–50% 40–60% 50–70% a a a Slope using V −2.02 (0.52) −1.76 (0.31) −1.97 (0.46) mean Slope using V −3.85 (0.42) −2.93 (0.71) −2.81 (0.63) peak Load0 using V (% of 1RM) 91.9% (15.3) 99.6% (14.9) 103.1% (14.0) mean Load0 using V (% of 1RM) 89.6% (7.7) 107.8% (23.8) 106.6% (10.2) peak a a a V 0 using V (m/s) 1.98 (0.22) 1.84 (0.15) 1.97 (0.31) mean mean a a V 0 using V (m/s) 3.47 (0.25) 3.05 (0.41) 2.97 (0.41) mean peak Indicates values differs significantly (P < 0.01) from V at that load range peak Indicates values differ significantly from the actual 1RM at that load range loads (30, 40 and 50% of 1RM) resulted in the most robust for the light relative loads, a rough estimate of accurate prediction of 1RM bench press strength (Fig. 4), the 1RM appears to be sufficient for this method. although there was a constant fixed bias towards under Although a fixed bias exists to under predict 1RM by estimating 1RM by approximately 9 kg. approximately 9 kg with using these loads, the accuracy of the model to predict Smith Machine bench press Discussion 1RM when using V during BP-T is quite high. Add- peak This study used the well-established linear-regression itionally, the high precision of this regression model is at techniques proposed by Jidovtseff and coworkers [16]to least comparable to other established prediction proce- examined which combination of three relative submaxi- dures that use more time-consuming protocols and/or mal loads during BP-T testing results in the best predic- also possibly have a greater potential for injury or soreness tion of 1RM Smith Machine bench press strength. We [5, 7–9, 12–15]. From a practical perspective, our findings also examined whether regression models developed suggest that there is no need to test over heavy relative using V , rather than the variable suggested by these and absolute loads [5, 6] when using the force-load tech- peak researchers (V ), provide the best prediction of bench nique to estimate Smith Machine 1RM in recreational and mean press 1RM. Finally, we examined whether performing a novice level weight trainers [5, 7, 8, 12, 14]. BP-T (instead of an explosive bench press) influences The finding that bar velocity data decreases with the nature of the regression model when using either of increasing relative load is not unique and simply con- these bar velocity measures. firms the standard exponential force velocity profile This study builds upon the findings of Gracia-Ramos first developed by Hill [33] nearly 80 years ago. Our and co-workers [24, 25], highlighting that 1RM bench data for V does however contain an anomaly at mean press strength on a Smith Machine can be determined 50% of 1RM (Fig. 2), suggesting that the V may be mean with acceptable levels of precision by examining V too gross a measure to be able to detect known mean and V during Smith Machine BP-T over three sub- changes in performance that occur across our load peak maximal loads. Perhaps even more importantly, our data ranges. Similarly, our results also suggest that V is peak suggests that the best and most reliable prediction amoreeffectivemeasurethan V when using this mean model was based on relative loads representing just 30, technique to predict 1RM in regular (but non-athletic) 40 and 50% of 1RM. Importantly, as the prediction is weight trainers with a mean 1RM Smith Machine bench press approximately equivalent to 1 body Table 2 CV% values of the slope, abscissa (Load0) and ordinate weight. While these findings are agreement with earl- (V 0) intercept data for each the regression lines developed mean ier research [20, 25]other studies favour V [24, 26], mean using both V and V across the three loading ranges mean peak highlighting that the specific loading regime influences Variable Percent of 1RM this outcome. 30–50% 40–60% 50–70% Slope using V 25.7 17.6 23.4 mean Slope using V 10.9 24.2 22.4 peak Table 3 ICC measured versus predicted 1RM Load0 using V (% of 1RM) 16.6 15.0 13.6 mean Variable Percent of 1RM Load0 using V (% of 1RM) 8.6 22.1 9.6 peak 30–50% 40–60% 50–70% V 0 using V (m/s) 11.1 8.2 15.7 V 0.868 (0.558–0.966) 0.855 (0.521–0.962) 0.849 (0.506–0.960) mean mean mean V 0 using V (m/s) 7.2 13.4 13.8 V 0.967 (0.890–0.990) 0.680 (0.204–0.896) 0.867 (0.604–0.960) mean peak peak Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 6 of 8 Fig. 3 The top row represents the three models to predict 1RM Smith Machine bench press based on mean vertical lifting velocity (V ). The mean left models are for the loads representing 30–50% of 1RM (●), the middle models for loads 40–60% of 1RM (♦) and the right models for loads representing 50–70% of 1RM (▲). The second row represents the respective Bland-Altman plots for each loading group Our V data are similar to data from physically ac- We acknowledge that our testing was based on a peak tive collegiate men performing a similar BP-T task [21], relatively small sample of a diverse but specific with values between studies differing by less than population of beginning weight training adults, how- 0.07 m/s at similar loads. Conversely the slope and ever these samples sizes are relatively common in V 0 data from our models using V to predict this domain. Additionally, our sample characteristics mean mean 1RM differ considerably from values from the original are typical for many healthy individuals who attend research using this method [16]. No doubt this is a func- health clubs and/or commercial strength training tion of the protocols adopted by these researchers that facilities. We also acknowledge that it the accuracy prevented the participants from releasing the bar (hence of regression models that attempt to predict values performing a dynamic bench press and not a BP-T per outside of the range of the collected data is severely se). The use of this approach by these researchers is compromised. However, this process is fundamental somewhat surprising as their V data would have to all research in this domain and so largely un- mean been effected by a pronounced bar deceleration near the avoidable. Importantly, we have not suggested that end of the movement [23] and so the accuracy of these the protocols presented in this project offer an exact data could be optimised. Importantly, V during BP-T estimate of a participant’s Smith Machine 1RM peak testing is not only reliable (CV% values between 1.7 and bench press. 3.3), but also presents with smaller CV% values than for dynamic bench press movements [21, 25]. From a prac- Conclusions tical stand point V is relatively simple to quantify, as Our results suggest that within this target population peak it can be measured using inexpensive devices (e.g. op- reliable estimates of Smith Machine 1RM bench press tical encoders or linear position transducers), or esti- strength can be achieved using the load-velocity ap- mated using bar throw height. These approaches can be proach with BP-T loads between 30 and 50% of 1RM. adopted easily in health clubs or commercial gymna- We do however acknowledge that the reliability and siums and provide acceptable predictions of 1RM that accuracy of the velocity based method presented here can be used in the development of more effective train- can suffer from fatigue or lack of motivation of the ing programs. athletes. However, issues such as these are systemic in Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 7 of 8 Fig. 4 The top row represents the three models to predict 1RM Smith Machine bench press based on peak vertical lifting velocity (V ). The left peak models are for the loads representing 30–50% of 1RM (●), the middle models for loads 40–60% of 1RM (♦) and the right models for loads representing 50–70% of 1RM (▲). The second row represents the respective Bland-Altman plots for each loading group nearly all strength assessment protocols and can be apply. Importantly, our results show that the most ac- managed with appropriate testing regimens. We also curate and reliable models are created from BP-T V peak acknowledge that these data are specific to Smith data (not V ), a variable that developed with min- mean Machine bench press and BP-T, and may not be imal post-testing processing, from loads representing transferable to conventional free weight testing. Fu- just 30, 40 and 50% of 1RM. Using the approach de- ture research should attempt to confirm these results scribed in our study exercise programmers can predict withalargersampleof participantsand conductap- 1RM Smith Machine bench press strength and monitor propriate between session reliability assessments. Add- performance enhancement with acceptable accuracy itionally, it would also be appropriate prospectively without the need to expose their clients to extremely compare the incidence of soreness and injury between heavy loads, or lift to fatigue protocols. the methods proposed in this study and traditional 1RM determination. Abbreviations 1RM: One repetition maximum; BM: Body mass; BP-T: Bench press throw; CV%: Coefficient of variance; ICC, 3,1: Intra class correlations; Load0: The Practical applications abscissa from the linear regression equation derived from load and mean Conventional 1RM bench press strength testing ex- vertical propulsive lifting velocity; LPT: Linear position transducer; Slope: The poses people to very high relative loads. Our findings gradient of the linear regression equation derived from load and mean vertical propulsive lifting velocity; V : Mean vertical propulsive velocity; mean indicate that in normal healthy populations bar velocity V 0: The ordinate from the linear regression equation derived from load mean data recorded during relatively light load Smith Ma- and mean vertical propulsive lifting velocity; V : Peak vertical lifting peak chine BP-T testing can be used to accurately predict velocity 1RM Smith Machine bench press strength. The large range in the submaximal load range allows practitioners Funding to estimate the 1RM as start point by using a team This project received no external funding. average, last season values or a weight dependent 1RM to define the submaximal test weights. Additionally, it Availability of data and materials is simple to determine V and V during BP-T mean peak The datasets used and/or analysed during the current study are available testing and the linear regression models are easy to from the corresponding author on reasonable request. Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 8 of 8 Authors’ contributions 14. Whisenant MJ, Panton LB, East WB, Broeder CE. Validation of submaximal MGLS was a major contributor in data analysis and writing the manuscript. prediction equations for the 1 repetition maximum bench press test on a MS collected and process the data and assisted with data analysis. MH group of collegiate football players. J Strength Cond Res. 2003;17(2):221–7. assisted with data processing and analysis. SL was a major contributor in 15. Clarkson PM, Hubal MJ. Exercise-induced muscle damage in humans. Am J data analysis and writing the manuscript. All authors read and approved the Phys Med Rehabil. 2002;81(11 Suppl):S52–69. final manuscript. 16. Jidovtseff B, Harris NK, Crielaard JM, Cronin JB. Using the load-velocity relationship for 1RM prediction. J Strength Cond Res. 2011;25(1):267–70. Ethics approval and consent to participate 17. Sugiura Y, Hatanaka Y, Arai T, Sakurai H, Kanada Y. Estimations of one This research was approved by the ETH Human Research Ethics Committee repetition maximum and isometric peak torque in knee extension based on (No. 2012-N-10). Participants were informed of the experimental procedures the relationship between force and velocity. J Strength Cond Res. 2016; and risks and provided their written informed consent prior to attending 30(4):980–8. several familiarisation sessions. 18. Meylan CM, Cronin JB, Oliver JL, Hughes MM, Jidovtseff B, Pinder S. The reliability of isoinertial force-velocity-power profiling and maximal strength Competing interests assessment in youth. Sports Biomech. 2015;14(1):68–80. The authors declare that they have no competing interests. 19. Jandacka D, Uchytil J. Optimal load maximizes the mean mechanical power output during upper extremity exercise in highly trained soccer players. J Strength Cond Res. 2011;25(10):2764–72. Publisher’sNote 20. Loturco I, Kobal R, Moraes JE, Kitamura K, Abad CC, Pereira LA, Nakamura FY. Springer Nature remains neutral with regard to jurisdictional claims in Predicting the maximum dynamic strength in bench press: the high published maps and institutional affiliations. precision of the bar velocity approach. J Strength Cond Res. 2017;31(4): 1127–31. Author details 21. Garcia-Ramos A, Padial P, Garcia-Ramos M, Conde-Pipo J, Arguelles- School of Health and Sport Sciences, University of the Sunshine Coast, Cienfuegos J, Stirn I, Feriche B. Reliability analysis of traditional and ballistic Maroochydore DC, QLD 4558, Australia. Institute for Biomechanics, ETH bench press exercises at different loads. J Hum Kinet. 2015;47:51–9. Zürich, Zürich, Switzerland. Swiss Federal Institute of Sport, Magglingen, 22. Banyard HG, Nosaka K, Haff GG. Reliability and validity of the load-velocity Switzerland. relationship to predict the 1RM back squat. J Strength Cond Res. 2017;31(7): 1897–904. Received: 19 May 2017 Accepted: 11 May 2018 23. Newton RU, Kraemer WJ, Hakkinen K, Humphries B, Murphy AJ. Kinematics, kinetics, and muscle activation during explosive upper body movements. J Appl Biomech. 1996;12(1):31–43. References 24. Garcia-Ramos A, Pestana-Melero FL, Perez-Castilla A, Rojas FJ, Haff GG. Mean 1. Fleck SJ, Kraemer W. Designing resistance training programs. 4th ed. velocity vs. mean propulsive velocity vs. peak velocity: which variable Champaign, IL: Human Kinetics; 2014. determines bench press relative load with higher reliability? J Strength 2. McMaster DT, Gill N, Cronin J, McGuigan M. A brief review of strength and Cond Res. 2018;32(5):1273–9. ballistic assessment methodologies in sport. Sports Med. 2014;44(5):603–23. 25. Garcia-Ramos A, Haff GG, Padial P, Feriche B. Reliability of power and 3. Levinger I, Goodman C, Hare DL, Jerums G, Toia D, Selig S. The reliability of velocity variables collected during the traditional and ballistic bench press the 1RM strength test for untrained middle-aged individuals. J Sci Med exercise. Sports Biomech. 2018;17(1):117–30. Sport. 2009;12(2):310–6. 26. Loturco I, Pereira LA, Cal Abad CC, Tabares F, Moraes JE, Kobal R, Kitamura 4. Kaelin ME, Swank AM, Adams KJ, Barnard KL, Berning JM, Green A. K, Nakamura FY. Bar velocities capable of optimising the muscle power in Cardiopulmonary responses, muscle soreness, and injury during the one strength-power exercises. J Sports Sci. 2017;35(8):734–41. repetition maximum assessment in pulmonary rehabilitation patients. J 27. Kraemer WJ, Ratamess NA, Fry AC, French DN. Strength testing: Cardiopulm Rehabil Prev. 1999;19(6):366–72. development and evaluation of methodology. In: Maud PJ, Foster CF, 5. Niewiadomski W, Laskowska D, Gasiorowska A, Cybulski G, Strasz A, Langfort editors. Physiological Assessment of Human Fitness. 2nd ed. Champaign, IL: J. Determination and prediction of one repetition maximum (1RM): safety Human Kinetics; 2006. p. 119–50. considerations. J Hum Kinet. 2008;19:109–19. 28. Argus CK, Gill ND, Keogh JWL, Hopkins WG. Assessing the variation in the 6. Rontu JP, Hannula MI, Leskinen S, Linnamo V, Salmi JA. One-repetition load that produces maximal upper-body power. J Strength Cond Res. 2014; maximum bench press performance estimated with a new accelerometer 28(1):240–4. method. J Strength Cond Res. 2010;24(8):2018–25. 29. Argus CK, Gill ND, Keogh JWL, Hopkins WG, Beaven CM. Changes in 7. Reynolds JM, Gordon TJ, Robergs RA. Prediction of one repetition maximum strength, power, and steroid hormones during a professional rugby union strength from multiple repetition maximum testing and anthropometry. J competition. J Strength Cond Res. 2009;23(5):1583–92. Strength Cond Res. 2006;20(3):584–92. 30. Pallarés JG, Sánchez-Medina L, Pérez CE, De La Cruz-Sánchez E, Mora- 8. Mayhew JL, Johnson BD, LaMonte MJ, Lauber D, Kemmler W. Accuracy of Rodriguez R. Imposing a pause between the eccentric and concentric prediction equations for determining one repetition maximum bench press phases increases the reliability of isoinertial strength assessments. J Sports in women before and after resistance training. J Strength Cond Res. 2008; Sci. 2014;32(12):1165–75. 22(5):1570–7. 31. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for 9. Kim PS, Mayhew JL, Peterson DF. A modified YMCA bench press test as a studies in sports medicine and exercise science. Med Sci Sports Exerc. 2009; predictor of 1 repetition maximum bench press strength. J Strength Cond 41(1):3–13. Res. 2002;16(3):440–5. 32. Haff GG, Ruben RP, Lider J, Twine C, Cormie P. A comparison of methods 10. Bazuelo-Ruiz B, Padial P, Garcia-Ramos A, Morales-Artacho AJ, Miranda MT, for determining the rate of force development during isometric midthigh Feriche B. Predicting maximal dynamic strength from the load-velocity clean pulls. J Strength Cond Res. 2015;29(2):386–95. relationship in squat exercise. J Strength Cond Res. 2015;29(7):1999–2005. 33. Hill AV. The heat of shortening and the dynamic constants of muscle. Proc 11. Picerno P, Iannetta D, Comotto S, Donati M, Pecoraro F, Zok M, Tollis G, R Soc Ser B. 1938;126(843):136–95. Figura M, Varalda C, Di Muzio D, et al. 1RM prediction: a novel methodology based on the force-velocity and load-velocity relationships. Eur J Appl Physiol Occup Physiol. 2016;116(10):2035–43. 12. 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. 2002;42(3):304–8. 13. Mayhew JL, Prinster JL, Ware JS, Zimmer DL, Arabas JR, Bemben MG. Muscular endurance repetitions to predict bench press strength in men of different training levels. J Sport Med Phys Fitness. 1995;35(2):108–13. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Sports Science, Medicine and Rehabilitation Springer Journals

The impact of test loads on the accuracy of 1RM prediction using the load-velocity relationship

Free
8 pages
Loading next page...
 
/lp/springer_journal/the-impact-of-test-loads-on-the-accuracy-of-1rm-prediction-using-the-DcoTdUFvaL
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Sports Medicine; Orthopedics; Rehabilitation Medicine
eISSN
2052-1847
D.O.I.
10.1186/s13102-018-0099-z
Publisher site
See Article on Publisher Site

Abstract

Background: Numerous methods have been proposed that use submaximal loads to predict one repetition maximum (1RM). One common method applies standard linear regression equations to load and average vertical lifting velocity (V ) data developed during squat jumps or three bench press throw (BP-T). The main aim of this mean project was to determine which combination of three submaximal loads during BP-T result in the most accurate prediction of 1RM Smith Machine bench press strength in healthy individuals. Methods: In this study combinations of three BP-T loads were used to predict 1RM Smith Machine bench press strength. Additionally, we examined whether regression models developed using peak vertical bar velocity (V ), peak rather than V , provide the most accurate prediction of Smith Machine bench press 1RM. 1RM Smith Machine mean bench press strength was measured directly in 12 healthy regular weight trainers (body mass = 80.8 ± 5.7 kg). Two to three days later a linear position transducer attached to the collars on a Smith Machine was used to record V and V during BP-T between 30 and 70% of 1RM (10% increments). mean peak Results: Repeated measures analysis of variance testing showed that the mean values for slope and ordinate intercept for the regression models at each of the load ranges differed significantly depending on whether V or mean V were used in the prediction models (P < 0.001). Conversely, the abscissa intercept did not differ significantly peak between either measure of vertical bar velocity at each load range. The key finding in this study was that 1RM Smith Machine bench press strength can be determined with high relative accuracy by examining V and V mean peak during BP-T over three loads, with the most precise models using V during loads representing 30, 40 and 50% peak of 1RM (R = 0.96, SSE = 4.2 kg). Conclusions: These preliminary findings indicate that exercise programmers working with normal healthy populations can accurately predict Smith Machine 1RM bench press strength using relatively light load Smith Machine BP-T testing, avoiding the need to expose their clients to potentially injurious loads. Keywords: Strength assessment, Dynamic strength, Predictive models, Bench press throws Background issues as it is not only time consuming, but the outcome is The quantification of the maximum load that can be lifted effected by factors such as athlete experience, technique, through a full range of motion, or one repetition maximum fatigue and motivation [2]. Traditional 1RM testing is con- (1RM), is fundamental to the design of resistance training sidered to be safe when it is conducted in appropriate set- programs [1]. Typically, 1RM is either measured directly or tings under the supervision of qualified practitioners [3, 4]. calculated indirectly using predictive models. The direct Regardless, this 1RM exposes athletes to large musculoskel- determination of 1RM suffers from a number of pragmatic etal forces, and there is some evidence that 1RM testing can be potentially injurious [5, 6]and mayalsobeimpracti- cal with novices and/or in clinical settings [7]. * Correspondence: msayers@usc.edu.au Indirect methods for 1RM quantification tend to School of Health and Sport Sciences, University of the Sunshine Coast, follow two different protocols, both of which rely on the Maroochydore DC, QLD 4558, Australia Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 2 of 8 use of linear regression modelling. The most common technology that, due to large reductions in pricing, is indirect protocols involve lifting submaximal loads to becoming increasing accessible to strength coaches. Re- failure [7–9], a procedure that is relatively common in searchers report high correlations using this methods trained athletes [10] and in agreement with the ACSM between Load0 and 1RM bench press (r = 0.98, n = 112, guidelines of 8–12 repetition that is often used in a clin- SEE = 4 kg [7%]), although the strength of this relation- ical setting, but rare in everyday activities. Although ship is no doubt influenced by the large range in relative relatively easy to administer, the accuracy of these ‘lift to loads assessed (30 to 95% of 1RM) [16, 22]. Additionally, failure’ models is also influenced by elements such as participants in the study by Jidovtseff and co-workers age, training experience, motivation and lifting tempo [16] were required to always hold the bar (i.e. prevented [7–9, 11]. The prediction of 1RM using these methods from performing a BP-T), which will have a marked ef- appears to be more accurate when heavier loads are used fect on V due to the deceleration of the bar near the mean [12–14], with the optimal number of repetitions for end of the lift [23]. Nevertheless, some questions remain these prediction models being less than 10 [8]. Accord- as to whether V or V provides superior predict- mean peak ingly, these protocols potentially suffer from the same ive measures. Recently, Gracia-Ramos and co-workers limitations associated with 1RM testing, with the need [24] report that V during Smith Machine bench mean to lift high relative loads whilst fatigued. Additionally, press is a superior predictor of 1RM when compared to these lift to failure protocols are also likely to generate V . However, these findings appear to be specific to peak post exercise muscle soreness in novices [15], potentially the testing protocols as these researchers highlight in dissuading them from future exercise participation. another study that V during bench press throws is peak Alternative indirect methods rely on the load-velocity the superior predictor of 1RM [25]. Regardless, V mean [6, 16] or force-velocity [10, 11, 17–19] relationships and and V appear to be greater predictors of optimal load peak linear or quasi linear models to predict 1RM from a for power training than traditional methods that advo- series of maximal effort lifts with submaximal loads. cate percentages of 1RM [26]. These protocols use either isoinertial sensors or linear Nevertheless, the question remains as to the efficacy of position transducers that are attached to the collars or the procedures proposed by Jidovtseff and co-workers bar of training devices like Smith Machines to record [16], particularly when testing novice or inexperienced force, average and/or peak vertical lifting velocity data weight trainers for which higher lifting loads may be from the concentric phase of movements like jump contraindicated. Therefore, it is important to determine squats or bench press throws (BP-T). Although these whether such high relative loads are required during movements are more common in high performance these submaximal test protocols (i.e. up to 95% of 1RM) training programs, the use of a Smith Machine and ap- and which combination of relative loads result in the propriately trained “Spotters” means that these exercises most accurate predictive model of 1RM bench press can be completed safely with novice participants (NB: strength. Accordingly, the purpose of this study was to some Smith Machines contain a pneumatic brake which use the prediction model developed by Jidovtseff and prevents the bar from descending rapidly – hence im- coworkers [16] to determine which combination of three proving exercise safety). Although 1RM data recorded submaximal loads during BP-T result in the most accur- on Smith Machines are typically 10% higher than those ate prediction of 1RM Smith Machine bench press recorded using free weights, there are no significant strength in healthy individuals. We also examined differences between predicted 1RM values when using whether the ability to release the bar during the BP-T these devices [20]. These protocols also have the advan- changes the nature of the prediction model. Addition- tage of being relatively quick to perform as they involve ally, we examined whether V or V provides a mean peak loads between 30 and 80% of 1RM [11, 16, 21] being better prediction of Smith Machine bench press 1RM lifted as rapidly as possible for only 2–4 repetitions. strength in these participants. Accordingly, the overall loading in these protocols is less than ‘lift to failure’ protocols and so the risk of injury Methods may be decreased, particularly when applied to relatively Experimental approach to the problem untrained populations [16]. To determine which combination of three loads during Arguably, the simplest of the load-velocity models [16] BP-T results in the most accurate prediction of 1RM applies standard linear regression equations to load and bench press strength we tested 12 healthy, regular mean vertical propulsive lifting velocity (V ) data weight trainers on two separate occasions. On the first mean from three different loads to develop slope, abscissa occasion 1RM bench press strength was recorded using (Load0) and ordinate (V 0) intercept data. Import- standard procedures and recorded to the nearest 1 kg mean antly, variables such as V or peak vertical velocity [27]. During the second data collection (2–3 days after mean (V ) can be measured using relatively inexpensive the first testing session) participants performed three peak Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 3 of 8 repetitions of BP-T at loads representing 30, 40, 50, 60 [2]. The bar handle was attached permanently to this and 70% of their 1RM. We then processed these braking system, resulting in a total weight of 23 kg. To load-velocity BP-T data using the techniques proposed record the vertical position of the bar a linear position by Jidovtseff and coworkers [16] to determine which transducer (LPT) (WS17KT, ASM, Moosinning, Germany) three load range (30–50% of 1RM, 40–60% of 1RM, was installed on the Smith Machine’s pneumatic brake, 50–70% of 1RM) resulted in the most accurate pre- with data subsequently sampled at 1000 Hz, A/D con- diction of 1RM bench press strength. We also exam- verted and stored on a computer, Subsequent data analysis ined whether BP-T V or V provides a more of the LPT measurement were performed in MATLAB, mean peak accurate prediction of bench press 1RM by comparing with velocity data developed from the raw LPT outputs each of the models developed using these variables. using the first central difference method. The BP-T testing was conducted in accordance with Participants well-established protocols [23, 28, 29] at loads represent- Participants for this study (n = 12) were all recreational ing 30, 40, 50, 60 and 70% relative to 1RM. The execu- weight trainers who had been weight training at least tion order was determined randomly using Microsoft twice a week for a minimum of 1 year (body mass (BM) = Excel in order to avoid possible order effects during the 80.8 ± 5.7 kg, 1RM 84 ± 18 kg, relative 1RM =1.04 BM [i.e. testing session. In order to minimise the effects of relative load is represented as a function of BM]). None of fatigue there were 2–4 min between repetitions with the participants were involved in heavy load strength three repetitions completed at each load. The eccentric training. Participants were informed of the experimental phase was at a self-chosen speed, with the participants procedures and risks and provided their written informed required to wait for the start signal before commencing consent prior to attending several familiarisation sessions. the concentric motion [30]. There was approximately at This research was approved by the institutional Human 2 s pause between the eccentric and concentric phases. Research Ethics Committee (No. 2012-N-10). The V , and V and maximum bar acceleration mean peak were calculated from the first and second differentials of Procedures the linear transducer data. These data were then used to All bench press and BP-T data were collected on a develop a linear regression model for the prediction of standard Smith Machine. This machine was modified 1RM [16] (Fig. 1). We subsequently developed slope, with a custom made magnetic braking system as a safety Load0 and V 0 data for each of these regressions mean mechanism. Once the bar was released this safety mech- models over each of the load ranges (i.e. 30–50% 1RM, anism prevented it from falling back on the participant 40–60% 1RM and 50–70% 1RM). Fig. 1 Sample data from one subject, three loads (solid circles) processed using standard load-velocity techniques [16]. Graph includes the regression line and the calculated peak mean vertical velocity (V 0), theoretical load at 0 m/s (Load0) and average vertical lifting velocity mean (V )at1RM mean Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 4 of 8 Statistical analyses Results The influence of load on the various bar kinematic vari- ANOVA testing indicated that V and V both mean peak ables were determined via a series of repeated measures showed large, significant reductions (P < 0.001, ES > 1.2) analysis of variance (ANOVA) tests. Post-hoc analyses for each respective increase in relative BP-T load, except were undertaken using paired t-Test with Bonferroni for V between 40 and 50% of 1RM (Fig. 2). Results mean corrections. Shapiro-Wilk and Mauchly’stest ofsphericity also showed that the mean values for slope and V 0 mean were applied during all ANOVA testing. Where data at each of the load ranges differed significantly depend- violated the sphericity assumption Greenhouse-Geisser ing on whether V or V were used in the predic- mean peak corrections were applied. The relative magnitude of tion models (Table 1). Conversely, Load0 data did not differences were quantified using standard Cohen’s Effect differ significantly between either bar velocity measures Size (ES) analyses, with the following descriptors used to at each load range. The CV% values range from 7.2 up define the relative magnitude of the ES: < 0.2 = trivial, to 27.5% (Table 2), with the ICC (Table 3) data show 0.2–0.6 = small,0.6–1.2 =medium/moderate,1.2–2.0 = excellent reliability for V at the lightest range weight peak large,and>2.0= very large [31]. The predictive accuracy whereas only moderate reliability was observed for the of the model developed by Jidovtseff and coworkers [16] weights between 40 and 60%. All other cases showed was assessed using the three lightest loads (30–50% 1RM), good reliability. Typically, greater levels of acceptable re- the three middle loads (40–60% 1RM) and the three liability [32] were recorded for V compared to V . peak mean heaviest loads (50–70% 1RM), with these data then com- There were no noticeable differences in any of the pared with the measured 1RM values. Bland-Altman plots models that used V to predict 1RM Smith Machine mean were used to assess whether there were any levels of bias bench press strength (R between 0.85–0.89). Similarly, in any of the models, with simple t-tests used to assess for there were no significant differences between predicted differences between the actual and predicted values. The and actual 1RM Smith Machine (P = 0.21 to 0.95) when coefficient of variance (CV%) and the intra class correla- using V , although the corresponding Bland-Altman mean tions (ICC, 3,1) for the predicted versus the measured plots highlighting some issues with the accuracy of these 1RM were also calculated. Statistical analysis were per- data (Fig. 3). Conversely, there were significant differ- formed using the statistics package SPSS for Windows ences between the predicted and actual 1RM Smith (version 20), with a confidence level of 95%. All data are Machine bench press values when using V at the peak presented at means ±1 standard deviation (SD) unless lightest of the load ranges to (P < 0.001). However, the stated otherwise. predicted 1RM values for V for these light relative peak Fig. 2 Mean (1SD) mean (V ) and peak bar (V ) vertical velocities at each of the relative loads. * Indicates data significantly different mean peak (P < 0.01) than the other loads Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 5 of 8 Table 1 Mean (±1SD) values of the slope, abscissa (Load0) and ordinate (V 0) intercept data for each regression line developed mean using both V and V across the three loading ranges mean peak Variable Percent of 1RM 30–50% 40–60% 50–70% a a a Slope using V −2.02 (0.52) −1.76 (0.31) −1.97 (0.46) mean Slope using V −3.85 (0.42) −2.93 (0.71) −2.81 (0.63) peak Load0 using V (% of 1RM) 91.9% (15.3) 99.6% (14.9) 103.1% (14.0) mean Load0 using V (% of 1RM) 89.6% (7.7) 107.8% (23.8) 106.6% (10.2) peak a a a V 0 using V (m/s) 1.98 (0.22) 1.84 (0.15) 1.97 (0.31) mean mean a a V 0 using V (m/s) 3.47 (0.25) 3.05 (0.41) 2.97 (0.41) mean peak Indicates values differs significantly (P < 0.01) from V at that load range peak Indicates values differ significantly from the actual 1RM at that load range loads (30, 40 and 50% of 1RM) resulted in the most robust for the light relative loads, a rough estimate of accurate prediction of 1RM bench press strength (Fig. 4), the 1RM appears to be sufficient for this method. although there was a constant fixed bias towards under Although a fixed bias exists to under predict 1RM by estimating 1RM by approximately 9 kg. approximately 9 kg with using these loads, the accuracy of the model to predict Smith Machine bench press Discussion 1RM when using V during BP-T is quite high. Add- peak This study used the well-established linear-regression itionally, the high precision of this regression model is at techniques proposed by Jidovtseff and coworkers [16]to least comparable to other established prediction proce- examined which combination of three relative submaxi- dures that use more time-consuming protocols and/or mal loads during BP-T testing results in the best predic- also possibly have a greater potential for injury or soreness tion of 1RM Smith Machine bench press strength. We [5, 7–9, 12–15]. From a practical perspective, our findings also examined whether regression models developed suggest that there is no need to test over heavy relative using V , rather than the variable suggested by these and absolute loads [5, 6] when using the force-load tech- peak researchers (V ), provide the best prediction of bench nique to estimate Smith Machine 1RM in recreational and mean press 1RM. Finally, we examined whether performing a novice level weight trainers [5, 7, 8, 12, 14]. BP-T (instead of an explosive bench press) influences The finding that bar velocity data decreases with the nature of the regression model when using either of increasing relative load is not unique and simply con- these bar velocity measures. firms the standard exponential force velocity profile This study builds upon the findings of Gracia-Ramos first developed by Hill [33] nearly 80 years ago. Our and co-workers [24, 25], highlighting that 1RM bench data for V does however contain an anomaly at mean press strength on a Smith Machine can be determined 50% of 1RM (Fig. 2), suggesting that the V may be mean with acceptable levels of precision by examining V too gross a measure to be able to detect known mean and V during Smith Machine BP-T over three sub- changes in performance that occur across our load peak maximal loads. Perhaps even more importantly, our data ranges. Similarly, our results also suggest that V is peak suggests that the best and most reliable prediction amoreeffectivemeasurethan V when using this mean model was based on relative loads representing just 30, technique to predict 1RM in regular (but non-athletic) 40 and 50% of 1RM. Importantly, as the prediction is weight trainers with a mean 1RM Smith Machine bench press approximately equivalent to 1 body Table 2 CV% values of the slope, abscissa (Load0) and ordinate weight. While these findings are agreement with earl- (V 0) intercept data for each the regression lines developed mean ier research [20, 25]other studies favour V [24, 26], mean using both V and V across the three loading ranges mean peak highlighting that the specific loading regime influences Variable Percent of 1RM this outcome. 30–50% 40–60% 50–70% Slope using V 25.7 17.6 23.4 mean Slope using V 10.9 24.2 22.4 peak Table 3 ICC measured versus predicted 1RM Load0 using V (% of 1RM) 16.6 15.0 13.6 mean Variable Percent of 1RM Load0 using V (% of 1RM) 8.6 22.1 9.6 peak 30–50% 40–60% 50–70% V 0 using V (m/s) 11.1 8.2 15.7 V 0.868 (0.558–0.966) 0.855 (0.521–0.962) 0.849 (0.506–0.960) mean mean mean V 0 using V (m/s) 7.2 13.4 13.8 V 0.967 (0.890–0.990) 0.680 (0.204–0.896) 0.867 (0.604–0.960) mean peak peak Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 6 of 8 Fig. 3 The top row represents the three models to predict 1RM Smith Machine bench press based on mean vertical lifting velocity (V ). The mean left models are for the loads representing 30–50% of 1RM (●), the middle models for loads 40–60% of 1RM (♦) and the right models for loads representing 50–70% of 1RM (▲). The second row represents the respective Bland-Altman plots for each loading group Our V data are similar to data from physically ac- We acknowledge that our testing was based on a peak tive collegiate men performing a similar BP-T task [21], relatively small sample of a diverse but specific with values between studies differing by less than population of beginning weight training adults, how- 0.07 m/s at similar loads. Conversely the slope and ever these samples sizes are relatively common in V 0 data from our models using V to predict this domain. Additionally, our sample characteristics mean mean 1RM differ considerably from values from the original are typical for many healthy individuals who attend research using this method [16]. No doubt this is a func- health clubs and/or commercial strength training tion of the protocols adopted by these researchers that facilities. We also acknowledge that it the accuracy prevented the participants from releasing the bar (hence of regression models that attempt to predict values performing a dynamic bench press and not a BP-T per outside of the range of the collected data is severely se). The use of this approach by these researchers is compromised. However, this process is fundamental somewhat surprising as their V data would have to all research in this domain and so largely un- mean been effected by a pronounced bar deceleration near the avoidable. Importantly, we have not suggested that end of the movement [23] and so the accuracy of these the protocols presented in this project offer an exact data could be optimised. Importantly, V during BP-T estimate of a participant’s Smith Machine 1RM peak testing is not only reliable (CV% values between 1.7 and bench press. 3.3), but also presents with smaller CV% values than for dynamic bench press movements [21, 25]. From a prac- Conclusions tical stand point V is relatively simple to quantify, as Our results suggest that within this target population peak it can be measured using inexpensive devices (e.g. op- reliable estimates of Smith Machine 1RM bench press tical encoders or linear position transducers), or esti- strength can be achieved using the load-velocity ap- mated using bar throw height. These approaches can be proach with BP-T loads between 30 and 50% of 1RM. adopted easily in health clubs or commercial gymna- We do however acknowledge that the reliability and siums and provide acceptable predictions of 1RM that accuracy of the velocity based method presented here can be used in the development of more effective train- can suffer from fatigue or lack of motivation of the ing programs. athletes. However, issues such as these are systemic in Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 7 of 8 Fig. 4 The top row represents the three models to predict 1RM Smith Machine bench press based on peak vertical lifting velocity (V ). The left peak models are for the loads representing 30–50% of 1RM (●), the middle models for loads 40–60% of 1RM (♦) and the right models for loads representing 50–70% of 1RM (▲). The second row represents the respective Bland-Altman plots for each loading group nearly all strength assessment protocols and can be apply. Importantly, our results show that the most ac- managed with appropriate testing regimens. We also curate and reliable models are created from BP-T V peak acknowledge that these data are specific to Smith data (not V ), a variable that developed with min- mean Machine bench press and BP-T, and may not be imal post-testing processing, from loads representing transferable to conventional free weight testing. Fu- just 30, 40 and 50% of 1RM. Using the approach de- ture research should attempt to confirm these results scribed in our study exercise programmers can predict withalargersampleof participantsand conductap- 1RM Smith Machine bench press strength and monitor propriate between session reliability assessments. Add- performance enhancement with acceptable accuracy itionally, it would also be appropriate prospectively without the need to expose their clients to extremely compare the incidence of soreness and injury between heavy loads, or lift to fatigue protocols. the methods proposed in this study and traditional 1RM determination. Abbreviations 1RM: One repetition maximum; BM: Body mass; BP-T: Bench press throw; CV%: Coefficient of variance; ICC, 3,1: Intra class correlations; Load0: The Practical applications abscissa from the linear regression equation derived from load and mean Conventional 1RM bench press strength testing ex- vertical propulsive lifting velocity; LPT: Linear position transducer; Slope: The poses people to very high relative loads. Our findings gradient of the linear regression equation derived from load and mean vertical propulsive lifting velocity; V : Mean vertical propulsive velocity; mean indicate that in normal healthy populations bar velocity V 0: The ordinate from the linear regression equation derived from load mean data recorded during relatively light load Smith Ma- and mean vertical propulsive lifting velocity; V : Peak vertical lifting peak chine BP-T testing can be used to accurately predict velocity 1RM Smith Machine bench press strength. The large range in the submaximal load range allows practitioners Funding to estimate the 1RM as start point by using a team This project received no external funding. average, last season values or a weight dependent 1RM to define the submaximal test weights. Additionally, it Availability of data and materials is simple to determine V and V during BP-T mean peak The datasets used and/or analysed during the current study are available testing and the linear regression models are easy to from the corresponding author on reasonable request. Sayers et al. BMC Sports Science, Medicine and Rehabilitation (2018) 10:9 Page 8 of 8 Authors’ contributions 14. Whisenant MJ, Panton LB, East WB, Broeder CE. Validation of submaximal MGLS was a major contributor in data analysis and writing the manuscript. prediction equations for the 1 repetition maximum bench press test on a MS collected and process the data and assisted with data analysis. MH group of collegiate football players. J Strength Cond Res. 2003;17(2):221–7. assisted with data processing and analysis. SL was a major contributor in 15. Clarkson PM, Hubal MJ. Exercise-induced muscle damage in humans. Am J data analysis and writing the manuscript. All authors read and approved the Phys Med Rehabil. 2002;81(11 Suppl):S52–69. final manuscript. 16. Jidovtseff B, Harris NK, Crielaard JM, Cronin JB. Using the load-velocity relationship for 1RM prediction. J Strength Cond Res. 2011;25(1):267–70. Ethics approval and consent to participate 17. Sugiura Y, Hatanaka Y, Arai T, Sakurai H, Kanada Y. Estimations of one This research was approved by the ETH Human Research Ethics Committee repetition maximum and isometric peak torque in knee extension based on (No. 2012-N-10). Participants were informed of the experimental procedures the relationship between force and velocity. J Strength Cond Res. 2016; and risks and provided their written informed consent prior to attending 30(4):980–8. several familiarisation sessions. 18. Meylan CM, Cronin JB, Oliver JL, Hughes MM, Jidovtseff B, Pinder S. The reliability of isoinertial force-velocity-power profiling and maximal strength Competing interests assessment in youth. Sports Biomech. 2015;14(1):68–80. The authors declare that they have no competing interests. 19. Jandacka D, Uchytil J. Optimal load maximizes the mean mechanical power output during upper extremity exercise in highly trained soccer players. J Strength Cond Res. 2011;25(10):2764–72. Publisher’sNote 20. Loturco I, Kobal R, Moraes JE, Kitamura K, Abad CC, Pereira LA, Nakamura FY. Springer Nature remains neutral with regard to jurisdictional claims in Predicting the maximum dynamic strength in bench press: the high published maps and institutional affiliations. precision of the bar velocity approach. J Strength Cond Res. 2017;31(4): 1127–31. Author details 21. Garcia-Ramos A, Padial P, Garcia-Ramos M, Conde-Pipo J, Arguelles- School of Health and Sport Sciences, University of the Sunshine Coast, Cienfuegos J, Stirn I, Feriche B. Reliability analysis of traditional and ballistic Maroochydore DC, QLD 4558, Australia. Institute for Biomechanics, ETH bench press exercises at different loads. J Hum Kinet. 2015;47:51–9. Zürich, Zürich, Switzerland. Swiss Federal Institute of Sport, Magglingen, 22. Banyard HG, Nosaka K, Haff GG. Reliability and validity of the load-velocity Switzerland. relationship to predict the 1RM back squat. J Strength Cond Res. 2017;31(7): 1897–904. Received: 19 May 2017 Accepted: 11 May 2018 23. Newton RU, Kraemer WJ, Hakkinen K, Humphries B, Murphy AJ. Kinematics, kinetics, and muscle activation during explosive upper body movements. J Appl Biomech. 1996;12(1):31–43. References 24. Garcia-Ramos A, Pestana-Melero FL, Perez-Castilla A, Rojas FJ, Haff GG. Mean 1. Fleck SJ, Kraemer W. Designing resistance training programs. 4th ed. velocity vs. mean propulsive velocity vs. peak velocity: which variable Champaign, IL: Human Kinetics; 2014. determines bench press relative load with higher reliability? J Strength 2. McMaster DT, Gill N, Cronin J, McGuigan M. A brief review of strength and Cond Res. 2018;32(5):1273–9. ballistic assessment methodologies in sport. Sports Med. 2014;44(5):603–23. 25. Garcia-Ramos A, Haff GG, Padial P, Feriche B. Reliability of power and 3. Levinger I, Goodman C, Hare DL, Jerums G, Toia D, Selig S. The reliability of velocity variables collected during the traditional and ballistic bench press the 1RM strength test for untrained middle-aged individuals. J Sci Med exercise. Sports Biomech. 2018;17(1):117–30. Sport. 2009;12(2):310–6. 26. Loturco I, Pereira LA, Cal Abad CC, Tabares F, Moraes JE, Kobal R, Kitamura 4. Kaelin ME, Swank AM, Adams KJ, Barnard KL, Berning JM, Green A. K, Nakamura FY. Bar velocities capable of optimising the muscle power in Cardiopulmonary responses, muscle soreness, and injury during the one strength-power exercises. J Sports Sci. 2017;35(8):734–41. repetition maximum assessment in pulmonary rehabilitation patients. J 27. Kraemer WJ, Ratamess NA, Fry AC, French DN. Strength testing: Cardiopulm Rehabil Prev. 1999;19(6):366–72. development and evaluation of methodology. In: Maud PJ, Foster CF, 5. Niewiadomski W, Laskowska D, Gasiorowska A, Cybulski G, Strasz A, Langfort editors. Physiological Assessment of Human Fitness. 2nd ed. Champaign, IL: J. Determination and prediction of one repetition maximum (1RM): safety Human Kinetics; 2006. p. 119–50. considerations. J Hum Kinet. 2008;19:109–19. 28. Argus CK, Gill ND, Keogh JWL, Hopkins WG. Assessing the variation in the 6. Rontu JP, Hannula MI, Leskinen S, Linnamo V, Salmi JA. One-repetition load that produces maximal upper-body power. J Strength Cond Res. 2014; maximum bench press performance estimated with a new accelerometer 28(1):240–4. method. J Strength Cond Res. 2010;24(8):2018–25. 29. Argus CK, Gill ND, Keogh JWL, Hopkins WG, Beaven CM. Changes in 7. Reynolds JM, Gordon TJ, Robergs RA. Prediction of one repetition maximum strength, power, and steroid hormones during a professional rugby union strength from multiple repetition maximum testing and anthropometry. J competition. J Strength Cond Res. 2009;23(5):1583–92. Strength Cond Res. 2006;20(3):584–92. 30. Pallarés JG, Sánchez-Medina L, Pérez CE, De La Cruz-Sánchez E, Mora- 8. Mayhew JL, Johnson BD, LaMonte MJ, Lauber D, Kemmler W. Accuracy of Rodriguez R. Imposing a pause between the eccentric and concentric prediction equations for determining one repetition maximum bench press phases increases the reliability of isoinertial strength assessments. J Sports in women before and after resistance training. J Strength Cond Res. 2008; Sci. 2014;32(12):1165–75. 22(5):1570–7. 31. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for 9. Kim PS, Mayhew JL, Peterson DF. A modified YMCA bench press test as a studies in sports medicine and exercise science. Med Sci Sports Exerc. 2009; predictor of 1 repetition maximum bench press strength. J Strength Cond 41(1):3–13. Res. 2002;16(3):440–5. 32. Haff GG, Ruben RP, Lider J, Twine C, Cormie P. A comparison of methods 10. Bazuelo-Ruiz B, Padial P, Garcia-Ramos A, Morales-Artacho AJ, Miranda MT, for determining the rate of force development during isometric midthigh Feriche B. Predicting maximal dynamic strength from the load-velocity clean pulls. J Strength Cond Res. 2015;29(2):386–95. relationship in squat exercise. J Strength Cond Res. 2015;29(7):1999–2005. 33. Hill AV. The heat of shortening and the dynamic constants of muscle. Proc 11. Picerno P, Iannetta D, Comotto S, Donati M, Pecoraro F, Zok M, Tollis G, R Soc Ser B. 1938;126(843):136–95. Figura M, Varalda C, Di Muzio D, et al. 1RM prediction: a novel methodology based on the force-velocity and load-velocity relationships. Eur J Appl Physiol Occup Physiol. 2016;116(10):2035–43. 12. 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. 2002;42(3):304–8. 13. Mayhew JL, Prinster JL, Ware JS, Zimmer DL, Arabas JR, Bemben MG. Muscular endurance repetitions to predict bench press strength in men of different training levels. J Sport Med Phys Fitness. 1995;35(2):108–13.

Journal

BMC Sports Science, Medicine and RehabilitationSpringer Journals

Published: May 29, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off