Effects of velocity loss in the bench press exercise on strength gains, neuromuscular adaptations, and muscle hypertroph

Effects of velocity loss in the bench press exercise on strength gains, neuromuscular adaptations, and muscle hypertrophy Fernando Pareja-Blanco1,2; Julián Alcazar3,4; Pedro J Cornejo-Daza1; Juan SánchezValdepeñas1; Carlos Rodriguez-Lopez3,4; Javier Hidalgo-de Mora1; Miguel SánchezMoreno5; Beatriz Bachero-Mena5; Luis M. Alegre3,4; Manuel Ortega-Becerra1,2 1Physical Performance & Sports Research Center, Department of Sports and Computers Sciences, Universidad Pablo de Olavide, Seville, Spain 2Faculty of Sport Sciences, Department of Sports and Computers Sciences, Universidad Pablo de Olavide, Seville, Spain 3GENUD Toledo Research Group, Universidad de Castilla-La Mancha, Toledo, Spain 4CIBER of Frailty and Healthy Aging (CIBERFES), Madrid, Spain 5Department of Physical Education and Sports, University of Seville, Seville, Spain Abstrac Objective: This study aimed to compare the effects of four velocity-based training (VBT) programs in bench press (BP) between a wide range of velocity loss (VL) thresholds— 0% (VL0), 15% (VL15), 25% (VL25), and 50% (VL50)—on strength gains, neuromuscular adaptations, and muscle hypertrophy. Methods: Sixty-four resistance-trained young men were randomly assigned into four groups (VL0, VL15, VL25, and VL50) that differed in the VL allowed in each set. Subjects followed a VBT program for 8-weeks using the BP exercise. Before and after the VBT program the following tests were performed: (a) cross-sectional area (CSA) measurements of pectoralis major (PM) muscle; (b) maximal isometric test; (c) progressive loading test; and (d) fatigue test. Results: Significant group x time interactions were observed for CSA (P < .01) and peak root mean square in PM (peak RMS-PM, P < .05). VL50 showed significantly greater gains in CSA than VL0 (P < .05). Only the VL15 group showed significant increases in peak RMS-PM (P < .01). Moreover, only VL0 showed significant gains in the early rate of force development (RFD, P = .05), while VL25 and VL50 improved in the late RFD (P ≤ .01-.05). No significant group × time interactions were found for any of the dynamic strength variables analyzed, although all groups showed significant improvements in all these parameters. Conclusion: Higher VL thresholds allowed for a greater volume load which maximized muscle hypertrophy, whereas lower VL thresholds evoked positive neuromuscular-related adaptations. No significant differences were found between groups for strength gains, despite the wide differences in the total volume accumulated by each group. KEYWORDS fatigue, neural adaptations, resistance training, structural adaptations, training prescription, velocity-based training INTRODUCTION The effectiveness of resistance training (RT) for enhancing muscle strength and hypertrophy, movement velocity, power output, and muscular endurance is widely recognized.1 Adaptations in response to RT may differ according to the manipulation of several variables, including training frequency, type and order of exercises, loading magnitude, volume (ie number of sets and repetitions), and repetition velocity.2 Exercise intensity during RT has traditionally been determined as lifted load relative to the one-repetition maximum (%1RM) or using percentages of set and repetition combination maximums;3 while RT volume has frequently been prescribed according to a theoretical maximum number of repetitions (MNR) per set that can be performed against a given %1RM up to muscle failure.1,2 However, a relatively high interindividual variation has been reported for the MNR that can be completed under a given %1RM.4-6 This issue may lead athletes to train at different levels of effort, defined as the relationship between the repetitions performed and the MNR that could be performed.7 Velocity-based training (VBT) has emerged as an objective method for real-time monitoring and prescription of RT intensity and volume.7,8 Firstly, strong relationships (R2 = .94-.98) between %1RM and movement velocity have been reported for exercises conducted on Smith machine such as bench press (BP),8 prone bench pull,9 pull-up,10 and different squat variants.11,12 Notably, it has also been reported that these relationships are not affected by individual strength levels or training background.13,14 These strong relationships open up the possibility of prescribing exercise intensity on a daily basis by adjusting the absolute load (kg) to match the movement velocity associated with the %1RM that is scheduled for the training session.8 Afterward, it has been shown that individual load-velocity relationships could provide more accurate predictions of %1RM from barbell velocity than general equations.15,16 However, it should be noted that velocity-based predictions conducted on freeweight exercises are not always accurate.17,18 Secondly, using only the velocity attained at a given %1RM is not a proper method to predict the MNR that can be completed with such load.19 However, the velocity loss (VL) incurred within the set, calculated as the relative difference between the fastest repetition velocity and the last repetition velocity of the set,7 has shown strong correlations (R2 = .96) with the percentage of completed repetitions with respect to the MNR.5,20 Accordingly, VL can be used to accurately determine the percentage of the MNR that has been completed in the set.5 Therefore, VBT can be considered as an alternative method to monitoring RT intensity and volume by collecting repetition velocity. VBT programs have been demonstrated to induce physiological adaptations in muscle function and structure and physical performance, these adaptations being dependent on the VL threshold.21,22 Thus, several studies conducted on the lower limbs (squat exercise) showed that greater muscle hypertrophic responses can be observed with higher VL thresholds (40%),21,22 although these VL thresholds may also induce negative neuromuscular adaptations,21 while greater strength gains were reported when applying moderate VL thresholds (10%-20%).21-24 However, it remains to be established whether these respective findings can be extrapolated to one of the most commonly prescribed upper-body RT exercises (ie the BP exercise). Therefore, the aim of the present study was to compare the effects of four BP VBT programs under different VL thresholds (0% vs 15% vs 25% vs 50%) on strength gains, neuromuscular adaptations, and muscle hypertrophy. MATERIAL AND METHODS Subjects This study is an extension of our previous work (see ParejaBlanco et al21). However, no variables or training interventions described here have been reported previously. Sixty-four resistance-trained young men (mean ± SD: age = 24.1 ± 4.3 years, height = 175.0 ± 5.5 cm, body mass = 75.5 ± 9.7 kg, relative 1RM BP = 0.90 ± 0.21 kg body mass), with at least 1.5 years of RT experience in the BP exercise, volunteered to take part in this study. Subjects were randomly assigned to one of the four training groups differed in the VL allowed during the set: 0% (VL0) vs 15% (VL15) vs 25% (VL25) vs 50%(VL50). Two subjects dropped out during the course of the study for reasons not related to the training intervention, so the remaining subjects per group were: VL0 (n = 15), VL15 (n = 16), VL25 (n = 15) and VL50 (n = 16). All subjects were informed about the purpose and test procedures and signed a written informed consent form before participating. The present study was approved by the Research Ethics Committee of “Hospitales Universitarios Virgen MacarenaVirgen del Rocío” (Reference: 1547-N-19), in accordance with the Declaration of Helsinki. Study design An experimental research design was used to examine the effects of four RT programs that differed in the VL threshold during the set in the BP exercise: 0%, 15%, 25%, and 50%. Two sessions per week (48-72 hours apart) were performed over 8 weeks as part of a progressive RT program. Subjects were asked to abstain from other types of vigorous physical activity involving the upper body during the research period. Subjects were measured on two occasions: 72 hours before (Pre-training) and 72 hours after (Posttraining) the 8-week training intervention. A battery of tests was performed in two testing sessions (separated by 48 hours). The first testing session consisted of cross-sectional area (CSA) measurements of the pectoralis major (PM) muscle. In the second testing session, a battery of tests was performed as follows: (a) isometric BP test; (b) progressive loading BP test; and (c) fatigue test, which consisted of performing repetitions to failure with the 70% of 1RM attained at Pre-training in the BP exercise. Training compliance was 100% for all sessions. Sessions were performed at the same time of the day for each subject (±1 hour) and under similar environmental conditions (~20°C and 60% humidity) in a research laboratory under the direct supervision of the researchers. Subjects were motivated to give maximal effort with strong verbal encouragement during all the test and training sessions. Testing procedures Ultrasonography B-mode ultrasonography (MyLab 25, Esaote Biomedica), with a 50 mm, 5-12 MHz linear-array probe, was used to assess the CSA of the PM muscle. Before the collection of the images, subjects remained lying in the supine position for 15 minutes, with their arms resting at the sides of the trunk, and palms facing down. The CSA of the PM was recorded in the sagittal plane at an intermediate point between the sternum and the right areola mammae of the participants. This intermediate point was determined after an exploratory analysis as the most lateral point of the PM muscle while avoiding imaging of the pectoralis minor muscle. Then, a straight line was drawn with a pen over the participant's skin of from the most inferior (~5th rib) to the most superior (clavicle) portion of the PM muscle. Finally, the extended field of view mode was used to register three images of the CSA of the PM muscle. Ultrasound images were recorded and digitally analyzed (ImageJ 1.51j8, NIH) by the same operator, who was blinded to subject allocation. The CSA of the PM muscle was measured by surrounding the bounds (aponeuroses) of the muscle (Figure 1A). Two ultrasound images were initially assessed, and if the coefficient of variation (CV) exceeded 5%, the third image was analyzed. The average value from all the analyzed images was considered for further analysis. Consistency in measurement sites across testing days was achieved by recording the probe positions on a transparent acetate sheet and using easily identifiable infiltrations of fatty and connective tissues as landmarks. Test-retest reliability in a sub-group of 10 subjects evaluated 24 hours apart was: CV = 3.9%. Isometric bench press test This test consisted of performing two 5 seconds maximal isometric contractions in the BP exercise, with a 1 minute rest between trials. The test was performed on a Smith machine (Fitness Line, Peroga) with the subjects placed in the supine position on top of a bench (Bench Fitness Line, Peroga) that was fitted onto a 0.8 × 0.8 m dynamometric platform (FP-500, Ergotech). The feet were positioned on the bench to record the force applied by the force platform. Before the test, the isometric position was individually adjusted with the bar placed 1 cm above the participant's chest through height-adjustable movable supports (elbow joint angle ~40°, considering full extension as 180°). A pronated grip at a width self-selected by the subject (approximately 150% of the biacromial distance) was used. Individual bar height and width grip were recorded to be repeated at Post-training sessions. Subjects were instructed to push against the bar as fast and hard as possible after the cue “ready, set, go!” External forces were collected at a sampling rate of 1000 Hz. Raw force-time data were automatically processed (4th order low-pass Butterworth filter with no phase shift using a 200 Hz cutoff frequency) with specialized software (TForce System, Ergotech). The following variables were measured during each attempt: (a) maximal isometric force (MIF); (b) maximal rate of force development (RFDmax), which was calculated as the maximum slope in the force-time curve in 20 ms time intervals; and (c) the average tangential slope of the force-time curve obtained over different time intervals (50, 100, 150, 200, and 400 ms from the onset of force production; RFD0-50, RFD0-100, RFD0-150, RFD0-200, and RFD0-400, respectively). The average value of the two attempts for each variable was recorded for further analysis. The onset of the force signal was established at the point where the signal raised above 3 SDs the baseline signal. Test-retest CV values were: 5.7% and 17.7% for MIF and RFDmax, respectively. For RFD obtained over different time intervals, CV values were as follows: 21.1%, 14.3%, 10.3%, 9.5%, and 6.7% for RFD0-50, RFD0-100, RFD0-150; RFD0-200, and RFD0-400, respectively. EMG signal acquisition EMG signals were recorded continuously during the isometric test. Electrodes were placed over the PM and triceps brachii (TB) muscles of the right side according to surface EMG recommendations for non-invasive muscle evaluation.25 Electrode positions were recorded onto a transparent acetate as well as different anatomic references and skin moles as landmarks in order to replicate the electrode positions at Post-training. EMG signals were collected using a parallel bar, bipolar, surface electromyographic sensor wireless Trigno™ EMG system (interelectrode distance of 10 mm, common mode rejection ratio >80 dB, and bandwidth filter between 20 and 450 Hz ± 10%) (Delsys Inc). The baseline noise was <5 µV peak-to-peak, and sampling rate was 1926 Hz. Raw EMG data were stored in digital format using EMG works Acquisition software (Delsys Inc) and smoothed by root mean square (RMS) calculation using a moving window of 100 ms with an overlap of 99 ms From each isometric trial, the highest averaged (over a 500 ms window) RMS value (peak EMG) and the integrated EMG (iEMG) of the RMS (normalized to peak EMG)-time curve over different time intervals (50, 100, 150, 200, and 400 ms from the onset of EMG activity, iEMG0-50, iEMG0-100, iEMG0-150, iEMG0-200, and iEMG0-400, respectively) were calculated. The average value of each variable of the two maximal isometric contractions was recorded for analysis. The onset of the EMG signal was set at the point where the signal raised above 3 SDs the baseline signal. Test-retest CV values for peak EMG were 13.6% and 18.1% for PM and TB muscles, respectively. Progressive loading test The individual load-velocity relationships and 1RM load in the BP exercise were determined using a Smith machine with no counterweight mechanism (Multipower Fitness Line, Peroga) and a linear velocity transducer (T-Force System Ergotech). The participants’ position and grip were the same as reported during the isometric test. During each repetition, the subjects were required to perform the eccentric phase in a controlled manner and to maintain a static position for ~1 second at the end of this phase (ie bar resting on the chest) to minimize the contribution of the rebound effect and allow for more reproducible measurements.26 Then, the bar was lifted at maximal intended velocity upon hearing the command. Throwing the bar at the end of the concentric phase was not allowed. The initial load consisted of one set of 3 repetitions with a rigid plastic bar (bar weight <0.2 kg) to assess the maximal unloaded velocity (V0) in BP exercise. Then, a set of 3 repetitions with 20 kg was performed and the load was progressively increased in 10 kg increments until the attained MPV was ≤0.30 m/s. When the load was close to 0.3 m/s, the load was increased in smaller increments (5 down to 2.5 kg) to allow better adjustments. A total of 6.3 ± 1.3 increasing loads were used for each subject. Three repetitions were executed for light (>0.80 m/s), two for medium (0.80-0.60 m/s) and only one for heavy (<0.60 m/s) loads. Inter-set recovery periods ranged from 3 minutes (light) to 5 minutes (heavy loads). Warm-up consisted of two sets of 6 BP repetitions with 0.2 and 20 kg, respectively. Only the fastest repetition with each load was considered for subsequent analysis. The velocity measures used in this study correspond to the mean velocity of the propulsive phase of each repetition (ie mean propulsive velocity, MPV), defined as that portion of the concentric action during which the measured acceleration is greater than acceleration due to gravity (−9.81 m/ s2) (Sanchez-Medina, Perez, & Gonzalez-Badillo, 2010). In addition to 1RM strength and V0, three other variables were analyzed: (a) average MPV attained against all absolute loads common to Preand Posttraining (AV); (b) average MPV attained against absolute loads that were lifted faster than 0.8 m/s at Pre-training (AV > 0.8); and (c) average MPV attained against absolute loads that were lifted slower than 0.8 m/s at Pre-training (AV < 0.8). These variables were analyzed to examine the effects on the different parts of the load-velocity relationship (ie velocity developed against light vs heavy loads). Fatigue test This test was performed with the same absolute load (kg) at Preand Post-training measurements, which corresponded to 70% of 1RM attained at Pre-training. The execution technique and devices used were the same as described in the progressive loading test. Subjects were required to complete as many repetitions as possible until muscle failure, performing each repetition at maximum intended velocity. The following variables were used for analysis: (a) maximal number of repetitions to failure (FT-MNR); and (b) average MPV attained against the same number of repetitions to Pre-training and Post-training (FT-AV). For example, if one subject performed 10 MNR at Pre-training and 15 MNR at Post-training, we evaluated the average MPV over the first 10 repetitions in both tests. This enabled assessment of the changes in MPV corresponding to the MNR at Pre-training. Fatigue testing began 5 minutes after subjects finished the BP progressive loading test. Resistance training program The descriptive characteristics of the RT program are presented in Table 1. The technical execution was identical to that previously described in the Progressive loading test section (including similar width grip and pause between the eccentric and concentric phases). All groups performed each training session with the same relative intensity (from 70% to 85% 1RM), number of sets (3) and interset recovery (4 minutes) in the BP exercise. Relative loads were determined from the individual load-velocity relationship obtained from the progressive loading test for each subject (R2 = .996 ± .004). Therefore, the absolute load (kg) was individually adjusted according to the individual velocity (±0.03 m/s) associated with the %1RM that was set for that session. We used a range of 0.03 m/s since it has recently been shown that the smallest detectable change in MPV when using the T-Force System is 0.03 m/s in BP exercise.27 The four groups differed in the VL threshold allowed in each set (0% vs 15% vs 25% vs 50%). VL0 performed only one repetition per set in order to induce the least possible fatigue. The rest of the groups finished their sets when the corresponding targeted VL threshold was exceeded. All repetitions during all sessions were recorded using a linear velocity transducer (T-Force System, Ergotech). The warmup preceding each training session was standardized for all training groups, as follows: 5 minutes of jogging at a selfselected easy pace, a set of 6 BP repetitions with 20 kg, followed by 3 sets of 6, 4 and 3 repetitions with loads of 40%, 50%, and 60% 1RM, respectively, for sessions 1-5 (in which the training load was 70% 1RM). An additional set of two repetitions with 70% 1RM was added for sessions 6-14 (in which the training load was 75%-80% 1RM), and a final set of one repetition with 80% 1RM was added for sessions 15 and 16 (in which the training load was 85% 1RM). A 3 minutes rest between the warm-up sets was always used. The total volume load (ie sets × reps × %1RM) and average training intensity (ie (total volume load/total repetitions) were calculated.3 Statistical analysis Data are reported as mean ± SD. Test-retest absolute reliability was measured by the standard error of measurement (SEM), which was expressed in relative terms through CV. The SEM was calculated as the root mean square of the intrasubject total mean square. Normality and homoscedasticity were verified with the Shapiro-Wilk and Levene's tests, respectively. Data were analyzed using a 4 × 2 factorial ANOVA with Bonferroni's posthoc comparisons, using one between-group factor (VL0 vs VL15 vs VL25 vs VL50) and one within-group factor (Prevs Post-training). A one-way ANOVA with Bonferroni posthoc adjustments was performed to analyze differences between groups in the training variables analyzed. Statistical significance was established at the P ≤ .05 level. In addition, effect size (ES) values were calculated using Hedge's g on the pooled SD28 using a purpose-built spreadsheet. The rest of statistical analyses were performed using SPSS software version 20.0 (SPSS Inc). Figures were designed using SigmaPlot 12.0 (Systat Software Inc). RESULTS No significant differences between groups were observed at Pre-training for any of the variables analyzed. Muscle cross-sectional area A significant group × time interaction was observed for PM muscle CSA (P = .008, Figure 1B). After completing the RT program, all groups obtained significant increases in muscle CSA (P < .001-.01). However, VL50 showed significantly greater gains than VL0 (P = .04) and almost significantly higher gains than VL15 (P = .06). No significant differences were reported between VL25 and VL50 or between any of the other groups (all P > .05). Isometric test No significant group × time interactions were observed for the isometric strength parameters analyzed, but significant overall time effects were observed for MIF, RFD0150, RFD0-200 and RFD0-400 (P < .001-.05). All groups showed significant gains in MIF (P < .001-.05). RT-induced changes in RFD parameters for each group are depicted in Figure 2. The VL0 intervention induced significant gains in RFD0-50 (P = .05), and VL25 improved in RFD0-400 (P < .05), whereas VL50 showed significant increases in RFD0-150 and RFD0-400 (P ≤ .01-.05). With regard to neuromuscular adaptations, a significant group × time interaction was observed for peak EMG-PM (P = .03) and there was an almost significant group × time interaction (P = .06) for peak EMG-TB. Only the VL15 group showed significant increases in peak EMG-PM (P < .01), while no significant changes were observed for peak EMG-TB for any group (Table 2). For normalized values, VL15 showed a significant decrease for iEMG400ms-PM (P < .05). No significant changes were detected for the rest of the parameters analyzed. Progressive loading test and fatigue test Changes in the selected performance variables from Preto Post-training for each group are reported in Table 3. No significant group × time interactions were found for any of the dynamic strength variables analyzed, but significant overall time effects were observed for all these parameters. All groups showed significant gains in 1RM strength (P < .001). With regard to the changes in the load-velocity relationship, all groups showed significant increases in AV, AV < 0.8, and AV > 0.8 (P < .001). However, only the VL15 group showed a significant improvement (P = .03) in the maximal unloaded velocity (V0). With regard to muscular endurance performance, the four groups showed significant enhancements in the fatigue test (FT-MNR and FT-AV, P < .001). Training program The average of the fastest repetitions measured in each session, which represents the %1RM lifted in each training session, was similar for all groups (Table 1). Likewise, no significant differences were observed in the average training intensity actually performed during the training program by each group. The mean velocity (MPV all reps) attained during the training program became slower as the VL threshold increased (P < .05, Table 1). Furthermore, the training volume (ie total volume load, total number of repetitions performed, the repetitions per set, and the repetitions performed with each %1RM) was higher as the VL threshold increased (P < .05, Table 1). The repetitions performed in the different velocity ranges are shown in Figure 3A. Figure 3B shows the evolution of the estimated 1RM strength (expressed as percentage of the Pre-training values) in each training session for all groups, based on the individual load-velocity relationship (R2 = .996 ± .004) for each subject. DISCUSSION This is the first study to analyze the effects of four different BP VBT programs (0% vs 15% vs 25% vs 50%) on strength gains, neuromuscular adaptations, and muscle hypertrophy of resistance-trained men. The main finding of this study was that the VL threshold in the set was a determining factor in modulating the muscle hypertrophic and neuromuscular adaptations that occur during RT. Higher VL thresholds (ie VL50), which accumulated remarkably higher volume load (Table 1) by performing more fatiguing and slower repetitions (Figure 3A), resulted in more muscle hypertrophy than lower VL thresholds (ie VL0), and only moderate VL thresholds (ie VL15) showed an increase in maximal neuromuscular excitation. In contrast, no group × time interactions were observed for RT-induced adaptations in muscle strength and endurance. It should be noted that these similar adaptations in BP strength levels were accompanied by very large differences in the total volume accumulated by each group throughout the study period (Table 1). After an 8-week VBT program, the four VL groups significantly improved their maximal dynamic and isometric strength levels (ie 1RM and MIF) and load-velocity relationshiprelated adaptations (ie AV, AV < 0.8; and AV > 0.8). In agreement with our findings, previous VBT studies carried out using lower body exercises (ie squat) have shown that higher VL thresholds do not induce further 1RM strength gains than lower VL thresholds.21,22,24 Moreover, all groups showed improvements in muscular endurance performance (ie FT-MNR and FT-AV), although the VL25 and VL50 groups attained the highest ES values. Supporting this finding, Izquierdo et al29 observed greater BP muscular endurance following a training program to failure compared with non-failure training, although there were no differences in the squat exercise. It should be noted that the VL0 group showed the lowest percentage improvements in all physical performance parameters. Accordingly, a certain minimal VL threshold and training volume seems to be necessary to elicit strength gains, although once a certain VL threshold is achieved, performing additional repetitions does not seem to elicit further strength gains. In an attempt to monitor strength evolution throughout the 8-week VBT, the 1RM was estimated in each training session from the individual load-velocity relationships of each subject. In this regard, all groups showed similar improvements in 1RM during the training program (Figure 3B). This result is very relevant for those athletes who are required to attain high strength levels during the entire season, with competitions every weekend or even every 3-4 days. It has been shown that RT protocols with large VL thresholds require longer recovery times (up to 48 hours post-exercise), whereas low VL thresholds show faster rates of recovery.30,31 Therefore, including RT protocols with lower VL thresholds during in-season periods with congested calendars could produce significant strength adaptations while maximizing recovery, compared with higher VL thresholds that may be detrimental to muscle recovery. Although training volume increased as VL threshold increased, this volume increment was due to accumulated slow and fatiguing repetitions (Figure 3A). These extra repetitions performed in the higher VL protocols may explain the different structural adaptations observed in the different groups. Although all training protocols induced muscle hypertrophy, higher VL thresholds (ie VL25-50) showed a greater hypertrophic response than lower VL thresholds (ie VL0-15). In agreement with our findings, it has been postulated that high training volumes32 and levels of effort (ie reaching or approaching muscle failure) are important factors in maximizing muscle growth.33,34 In this regard, a previous VBT study using the squat exercise also showed higher hypertrophy in the vastus lateralis muscle following an RT with higher VL (20%-40%) compared with lower VL (0%-10%).21 The higher exercise-induced metabolic and mechanical stress,7 the greater secretion of growth-promoting hormones and muscle damage, along with the higher total volume load observed in RT protocols with high VL thresholds30 could be responsible for the greater hypertrophic adaptations observed with these protocols.34,35 In addition, increased ribosomal biogenesis has recently been found to be behind the higher hypertrophic adaptations observed in high-volume RT programs.36 In contrast, a lower VL threshold seems to be required to induce positive neuromuscular adaptations, since only VL15 showed enhanced peak PM muscle excitation. Neuromuscular adaptations may be related to increased motor unit firing frequency, increased incidence of discharge doublets, and changes in fiber types and sarcoplasmic reticulum calcium kinetics.37-39 In addition, only VL15 showed improved maximal unloaded (<0.2 kg) velocity (V0). A faster muscle fiber phenotype has been reported to FIGURE 1 A, Ultrasound image obtained from the pectoralis major muscle of a standard subject. The cross-sectional area of the vastus lateralis muscle is surrounded by the yellow line. CL, clavicle; PM, pectoralis major; R, rib; RA, rectus abdominis. B, Changes produced on pectoralis muscle thickness, illustrated using ultrasound images from Preto Post-training for each group. N = 62. VL0, group that trained with a mean velocity loss of 0% in each set (n = 15); VL15, group that trained with a mean velocity loss of 15% in each set (n = 16); VL25, group that trained with a mean velocity loss of 25% in each set (n = 15); VL50, group that trained with a mean velocity loss of 50% in each set (n = 16); ES, within-group effect size from preto post-training. Intragroup significant differences from Preto Posttraining: **P ≤ .01, ***P ≤ .001. Statistically significant differences with VL0 group: 0P ≤ .05. Significant group × time interaction: #P ≤ .05 FIGURE 2 A, Changes produced in maximal rate of force development (RFDmax) from Preto Post-training for each group. B, Changes produced in 0-50 ms rate of force development (RFD0-50) from Preto Post-training for each group. C, Changes produced in 0-100 ms rate of force development (RFD0-100) from Preto Post-training for each group. D, Changes produced in 0-150 ms rate of force development (RFD0-150) from Preto Post-training for each group. E, Changes produced in 0-200 ms rate of force development (RFD0-200) from Preto Post-training for each group. F, Changes produced in 0-400 ms rate of force development (RFD0-400) from Preto Post-training for each group. N = 62; VL0, group that trained with a mean velocity loss of 0% in each set (n = 15); VL15, group that trained with a mean velocity loss of 15% in each set (n = 16); VL25, group that trained with a mean velocity loss of 25% in each set (n = 15); VL50, group that trained with a mean velocity loss of 50% in each set(n = 16); ES, within-group effect size from Preto Post-training. Intragroup significant differences from Preto Posttraining: * P ≤ .05, ** P ≤ .01, *** P ≤ .001 FIGURE 3 A, Number of repetitions in the BP exercise performed in each velocity range, and total number of repetitions completed by four training groups. B, Evolution of the estimated 1RM strength in the bench press exercise in each training session expressed as a percentage of the initial Pre-training level for four experimental groups. Data are mean ± SD, N = 62. VL0, group that trained with a mean velocity loss of 0% in each set (n = 15); VL15, group that trained with a mean velocity loss of 15% in each set (n = 16); VL25, group that trained with a mean velocity loss of 25% in each set(n = 15); VL50, group that trained with a mean velocity loss of 50% in each set (n = 16). Statistically significant differences with VL0 group: 0P ≤ .05. Statistically significant differences with VL15 group: 15P ≤ .05. Statistically significant differences with VL25 group: 25P ≤ .05 TABLE 1 Descriptive characteristics of the 8-wk velocity-based bench press training program performed by the four experimental groups Scheduled Session 1 Session 2 Session 3 Session 4 Session 5 Session 6 Session 7 Session 8 Set × %1RM 3 × 70 3 × 70 3 × 70 3 × 70 3 × 70 3 × 75 3 × 75 3 × 75 Target velocity (m/s) 0.65 ± 0.07 0.65 ± 0.07 0.65 ± 0.07 0.65 ± 0.07 0.65 ± 0.07 0.57 ± 0.07 0.57 ± 0.07 0.57 ± 0.07 Scheduled Session 9 Session 10 Session 11 Session 12 Session 13 Session 14 Session 15 Session 16 Set × %1RM 3 × 75 3 × 75 3 × 80 3 × 80 3 × 80 3 × 80 3 × 85 3 × 85 Target velocity (m/s) 0.57 ± 0.07 0.57 ± 0.07 0.49 ± 0.06 0.49 ± 0.06 0.49 ± 0.06 0.49 ± 0.06 0.41 ± 0.05 0.41 ± 0.05 Actually performed Fastest MPV (m/s) Slowest MPV (m/s) MPV all reps (m/s) Mean VL (%) Total Rep Total volume load (set × rep × %1RM) VL0 0.53 ± 0.05 0.48 ± 0.05 15 25 50 0.51 ± 0.0550 0.0 ± 0.0 15 25 50 48.0 ± 0.0 15 25 50 3621 ± 82 15 25 50 VL15 0.57 ± 0.05 0.43 ± 0.0425 50 0.52 ± 0.0550 16.3 ± 0.825 50 136.6 ± 17.825 50 10 666 ± 2627 25 50 VL25 0.57 ± 0.07 0.38 ± 0.0550 0.50 ± 0.0750 25.0 ± 0.750 191.1 ± 34.150 14 595 ± 353650 VL50 0.55 ± 0.07 0.21 ± 0.02 0.40 ± 0.05 51.9 ± 1.6 316.4 ± 65.1 22 687 ± 5734 Actually Average rep per Rep per set with Rep per set with Rep per set with Rep per set with Average Training performed set in all sessions 70% 1RM 75% 1RM 80% 1RM 85% 1RM Intensity (%1RM) VL0 1.0 ± 0.0 15 25 50 1.0 ± 0.0 15 25 50 1.0 ± 0.0 15 25 50 1.0 ± 0.0 15 25 50 1.0 ± 0.0 15 25 50 75.2 ± 1.7 VL15 2.9 ± 0.425 50 3.4 ± 0.325 50 2.9 ± 0.525 50 2.5 ± 0.525 50 2.0 ± 0.450 74.3 ± 0.8 VL25 4.0 ± 0.750 5.0 ± 0.950 4.1 ± 0.750 3.4 ± 0.850 2.6 ± 0.750 74.3 ± 0.9 VL50 6.6 ± 1.4 8.3 ± 1.7 6.7 ± 1.4 5.5 ± 1.5 4.2 ± 1.2 74.3 ± 0.9 Note: Data are mean ± SD. Only one exercise (bench press) was used in training. Abbreviations: Average rep per set in all sessions, average number of repetitions performed in each set; Average training intensity, average relative intensity attained during the training program calculated as total volume load/total repetitions; Fastest MPV, average of the fastest repetitions measured in each session (this value represents the average intensity, %1RM, achieved during the training program); Mean velocity loss, average velocity loss attained during the entire training program; MPV all reps, average MPV attained during the entire training program; MPV, mean propulsive velocity; Rep per set with a given %1RM, average number of repetitions performed in each set with each of the loads used (70, 75, 80 or 85%1RM); Slowest MPV, average of the slowest repetitions measured in each session; Target velocity, velocity associated with the scheduled %1RM; Total rep, Total number of repetitions performed during the training program; Total volume load, sets × reps × %1RM; VL0, group that trained with a mean velocity loss of 0% in each set (n = 15); VL15, group that trained with a mean velocity loss of 15% in each set (n = 16); VL25, group that trained with a mean velocity loss of 25% in each set (n = 15); VL50, group that trained with a mean velocity loss of 50% in each set (n = 16); VL, magnitude of velocity loss expressed as percent loss in mean repetition velocity from the fastest (usually first) to the slowest (last one) repetition of each set. Statistically significant differences with VL15 protocol: 15P ≤ .05. Statistically significant differences with VL25 protocol: 25P ≤ .05. Statistically significant differences with VL50 protocol: 50P ≤ .05.