RSPB 1260767 17 November 2016
Initial
CE: ####### QA: SM Coll:XX QC:#######
SPORTS BIOMECHANICS, 2016 VOL. XX, NO. XX, 114 http://dx.doi.org/10.1080/14763141.2016.1260767
Reliability of power and velocity variables collected during the traditional and ballistic bench press exercise Amador García-Ramosa, G. Gregory Haffb, Paulino Padiala and Belén Ferichea 5
a
Faculty of Sport Sciences, Department of Physical Education and Sport, University of Granada, Granada, Spain; bCenter for Exercise and Sport Science Research, Edith Cowan University, Joondalup, Australia AQ2
ABSTRACT
10
15
20
25
This study aimed to examine the reliability of different power and velocity variables during the Smith machine bench press (BP) and bench press throw (BPT) exercises. Twenty-two healthy men conducted four testing sessions after a preliminary BP one-repetition maximum (1RM) test. In a counterbalanced order, participants performed two sessions of BP in one week and two sessions of BPT in another week. Mean propulsive power, peak power, mean propulsive velocity, and peak velocity at each tenth percentile (20–70% of 1RM) were recorded by a linear transducer. The within-participants coefficient of variation (CV) was higher for the load–power relationship compared to the load–velocity relationship in both the BP (5.3% vs. 4.1%; CV ratio = 1.29) and BPT (4.7% vs. 3.4%; CV ratio = 1.38). Mean propulsive variables showed lower reliability than peak variables in both the BP (5.4% vs. 4.0%, CV ratio = 1.35) and BPT (4.8% vs. 3.3%, CV ratio = 1.45). All variables were deemed reliable, with the peak velocity demonstrating the lowest within-participants CV. Based upon these findings, the peak velocity should be chosen for the accurate assessment of BP and BPT performance.
ARTICLE HISTORY
Received 23 February 2016 Accepted 24 October 2016 KEYWORDS
Load–power relationship; load–velocity relationship; linear position transducer; peak velocity
Introduction
30
35
AQ1
A common aim of many resistance-training programmes is to create a training adaptation that allows the athlete to produce more force with the same absolute load, or in other words, develop higher velocities while moving the same absolute load. In this context, the determination of the individual load–velocity or load–power relationships during multi-joint exercises is a useful tool for monitoring an athlete’s training status (Cormie, McGuigan, & Newton, 2010; González-Badillo, Rodríguez-Rosell, Sánchez-Medina, Gorostiaga, & Pareja-Blanco, 2014; Sanchez-Medina, Perez, & Gonzalez-Badillo, 2010). The use of loaded functional movements (e.g. bench press [BP], bench press throw [BPT], half-squat, etc.) to evaluate strength and power characteristics are justified by the close relationship that has been shown between these measures and sports performance (Baker, 2001; Gorostiaga, Granados, Ibáñez, & Izquierdo, 2005; Marques, van den Tillaar, Vescovi, & González-Badillo, 2007). CONTACT Amador García-Ramos
[email protected]
© 2016 Informa UK Limited, trading as Taylor & Francis Group
RSPB 1260767 17 November 2016
2
5
10
15
20
25
30
35
Initial
CE: ####### QA: SM Coll:XX QC:#######
A. GARCÍA-RAMOS ET AL.
The implementation of a linear position transducer as a device for establishing these relationships has recently increased in popularity in the scientific and applied literature (García-Ramos, Tomazin et al., 2016; Harris, Cronin, Taylor, Boris, & Sheppard, 2010; Pareja-Blanco et al., 2016). These devices are usually attached to the bar and used to estimate the mechanical variables (force, velocity, and power) using the displacement-time data collected during the analysed movement. The mean and peak values of velocity, force, and power output obtained at each load tested can then be used to determine the athletes’ performance profile. In addition, linear transducers provide real time feedback of the velocity achieved at each repetition. Based on this information, a new resistance training methodology is being developed in which movement velocity is used as the criterion for the intensity and volume of the training session (velocity-based resistance training) (GonzálezBadillo, Marques, & Sánchez-Medina, 2011; Randell, Cronin, Keogh, Gill, & Pedersen, 2011; Sánchez-Medina & González-Badillo, 2011). When these assessments are performed with traditional resistance training exercises, such as the BP, using light and medium loads, the athlete will activate their antagonist muscles to apply force in the opposite direction of the load’s motion in order to not release the bar, effectively creating a deceleration effect at the end of the movement (Jarić, Ropret, Kukolj, & Ilić, 1995). This point is usually identified as the moment in which the acceleration of the bar becomes lower than gravity (-9.81 m/s2) (Sanchez-Medina et al., 2010). In order to account for this fact, the use of mean propulsive values (considers only the phase of the concentric phase when acceleration of the bar is higher than gravity) may be a preferred analytical measure in order to not underestimate an athlete’s performance as traditional mean values consider the entire concentric phase and may be impacted by the deceleration effect that occurs at the end of the concentric movement (Sanchez-Medina et al., 2010). The BP and BPT are probably the two exercises most used to assess upper-body power and create a load-velocity performance profiles (Baker, 2001; González-Badillo et al., 2014; Sanchez-Medina et al., 2010). Many variables have been employed to determine an athlete’s performance capacity under different loading conditions. However, to the best of our knowledge, no previous study has analysed which variable present the highest reliability. Therefore, the main objective of the present study was to examine the reliability of different power and velocity variables measured by a linear position transducer in the BP and BPT exercises with loads corresponding to 20–70% of one-repetition maximum (1RM). We hypothesised that velocity variables would be more reliable than power variables because the manipulation of the raw displacement-time data needed to obtain power output can magnify the measurement error (McMaster, Gill, Cronin, & McGuigan, 2014). Additionally, peak values were hypothesised to be more reliable than mean propulsive values as they are less affected by arbitrary decisions about how to determine the exact start and end points of the concentric phase (García-Ramos, Štirn et al., 2016; Hori et al., 2007).
Methods 40
Participants Twenty-two physically active collegiate men (age 21.7 ± 3.0 years, body mass 74.9 ± 7.5 kg, height 176.3 ± 5.2 cm) participated in this study. The 1RM in the BP exercise was 82.9 ± 12.2 kg (relative BP: 1.11 ± 0.13 kg/kg body mass). All participants were informed
RSPB 1260767 17 November 2016
Initial
CE: ####### QA: SM Coll:XX QC:#######
SPORTS BIOMECHANICS
3
of the procedures to be utilised and signed a written informed consent form prior to investigation. The study protocol adhered to the tenets of the Declaration of Helsinki and was approved by the University of Granada Institutional Review Board. Experimental design 5
10
15
20
25
30
35
A repeated-measures design was used to compare the test–retest reliability for different power and velocity variables collected during the Smith machine BP and BPT exercises. After a preliminary testing session (body composition and maximal dynamic strength [1RM] in BP), participants reported to the laboratory on four separate occasions during two consecutive weeks. Two sessions were performed each week with 48 h between each session. Participants were then randomly assigned so half performed both sessions of BP in the first week and the two sessions of BPT in the second week. Conversely, the other group performed the two BPT session in week 1 and the two BP sessions in week 2. Individual load–power and load–velocity relationships at each tenth percentile (20–70% of 1RM) were then determined on each testing day with the use of a linear position transducer. Preliminary testing (session 1) Participants came to testing after refraining from strenuous exercise for a minimum of 48 h. As soon as they arrived, their height (Seca 202, Seca Ltd., Hamburg, Germany) and body mass (Tanita BC-418 MA, Tanita corporation, Tokyo, Japan) were assessed. Before the commencement of the 1RM test, participants self-selected their hand position, which was measured and kept constant throughout all testing sessions. The warm-up consisted of joint mobility, dynamic stretching, and two sets of five repetitions with 20 and 30 kg, respectively. Participants then completed an incremental loading test to determine the BP 1RM. Initial load was set at 40 kg for all participants, and was progressively increasing in steps of 10–1 kg until the heaviest load that could be properly lifted was performed. Participants performed 1–2 repetitions per load and the recovery time between attempts was at least five minutes. Participants initiated the BP exercise with their elbows fully extended and then lowered the barbell in a continuous motion in a time frame of 1.5 s until the bar position was 1–2 cm above their intermammary line. Once this position was achieved, the participants were required to maintain this position for 1 s (velocity = 0 m/s). From this position, each participant was instructed to perform a purely concentric action to regain the initial position. The duration of eccentric and isometric phases was administered by auditory feedback through an ad hoc audio file, while the concentric phase was always performed explosively, at the maximum possible speed. The only difference in the execution between BP and BPT was that during the BPT participants accelerated the bar during the entire range of movement with the intention of throwing it as high as possible. Conversely, in BP, the bar had to be voluntarily decelerated at the end of the range of motion in order to not throw it. Reliability testing (sessions 2–5)
40
Participants reported to the laboratory on four separate occasions during two consecutive weeks. In one week, participants performed the two BP sessions, and in the other week, they performed the two BPT sessions in a counterbalanced fashion. To compensate for a
RSPB 1260767 17 November 2016
4
5
10
15
20
25
Initial
CE: ####### QA: SM Coll:XX QC:#######
A. GARCÍA-RAMOS ET AL.
possible training effect, the order of the exercises was randomly assigned. Recovery time between testing sessions was 48 h between sessions of the same exercise and 120 h before the non-tested exercise was tested. Every session for the same participant was carried out at the same time of the day and under similar environmental conditions. Each testing session was preceded by a 10-min standardised warm-up, which included dynamic stretching, arm and shoulder mobilisation and one set of six repetitions performed in an explosive manner with an external load of 17 kg in the assessed exercise (BP or BPT). After completing the standardised warm-up, an incremental loading test at each tenth percentile (20–70% of 1RM) was performed. Participants performed three repetitions (as quickly as possible) with each load, but only the repetition with the highest peak velocity of each load was considered for subsequent analysis. Recovery time between sets was at least five minutes in order to ensure full recovery. All tests were performed in a Smith machine (Technogym, Barcelona, Spain) that was instrumented with a linear position transducer (Real Power Pro Globus, Codogne, Italy) that sampled at 1,000 Hz and was used to automatically calculate kinetic (power) and kinematic (velocity) variables for every repetition. Four variables were analysed in the present study: • Mean propulsive power: mean power value of the propulsive phase of each repetition. • Peak power: maximum instantaneous power value reached during the concentric phase at a given load. • Mean propulsive velocity: mean velocity value of the propulsive phase of each repetition. • Peak velocity: maximum instantaneous velocity value reached during the concentric phase at a given load. The propulsive phase was defined as that portion of the concentric phase during which the measured acceleration is greater than acceleration due to gravity (i.e. bar acceleration > -9.81 m/s2) (Sanchez-Medina et al., 2010). Statistical analyses
30
35
40
Paired t-tests were used to determine if there were significant differences in mean propulsive power, peak power, mean propulsive velocity, and peak velocity between the testing sessions 1 and 2 for each exercise separately (BP and BPT). The magnitude of between-session differences was also expressed as a standardised mean difference (Cohen’s d effect size; ES). The criteria to interpret the magnitude of the ES were as follows: 2 = very large (Hopkins, Marshall, Batterham, & Hanin, 2009) Relative reliability was calculated by determining the intraclass correlation coefficient (two-way random effect model; ICC2,1) (Stock, Beck, DeFreitas, & Dillon, 2011; Weir, 2005). Additionally, the standard error of measurement (SEM), calculated as the square root of the mean square error term from the ANOVA, was expressed as a percentage of the participants’ mean scores (coefficient of variation; CV) to assess absolute reliability (Hoefelmann et al., 2015; Weir, 2005). Acceptable reliability was determined as an ICC > 0.70 and a CV 0.70 and CV
