Acute Effect of the Bounce Squat on Ground Reaction Force at the Turning Point and Barbell Kinematics

Supplemental Digital Content is Available in the Text.

Abstract

Achermann, BB, Drewek, A, and Lorenzetti, SR. Acute effect of the bounce squat on ground reaction force at the turning point and barbell kinematics. J Strength Cond Res 40(1): 1–8, 2026—The free-weight back squat is a key exercise for developing lower-body strength, with variations that influence muscle activation and performance. The bounce squat, a variation involving a controlled bounce, increases the eccentric velocity at the squat's bottom. In this study, we aimed to investigate the effects of the bounce squat, descent velocity, and load on barbell kinematics and ground reaction force (GRF) during back squats. In addition, we explored the prediction of GRF from barbell kinematics. Overall, 29 participants completed 2 sessions, including bounce and no-bounce squats. Session 1 was focused on the load (70 and 80% 1RM), whereas session 2 examined the descent velocity (fast vs. slow). Data were analyzed using linear mixed-effects models (p = 0.05). Bounce squats increased the GRF significantly by 19 and 22% (sessions 1 and 2, respectively). Furthermore, it increased the peak velocity in the early concentric phase (Vpeak1) and reduced it in the late concentric phase (Vpeak2). Higher loads significantly reduced the mean concentric velocity (Vmean), minimum velocity, Vpeak1, and Vpeak2, and minimally enhanced the GRF. Faster descent enhanced Vmean and Vpeak2 without affecting the GRF or Vpeak1. These findings emphasize balancing performance with musculoskeletal stress and underscore the importance of accurate technical execution of back squats. Supporting factors such as descent velocity and load should be tailored to preserve the optimal technique. Predicting GRF from barbell kinematics highlights its potential for monitoring squat performance.

Introduction

Resistance training is used widely by athletes to improve neuromuscular abilities, including strength, power, and hypertrophy (34). Back squats are key exercises targeting the lower extremities (33). Furthermore, it is a multijoint lower-limb exercise that comprises hip, knee, and ankle exercises and helps improve lower-body strength. Variations in execution, such as foot placement, ankle height, squat depth, barbell placement, time under tension, and barbell load, result in different muscle activation patterns and serve varying purposes (3,10,14,15,17,20,27,30,32,36). Considering the widespread use of the free-weight back squat for athletic development, its role in different sports such as powerlifting (8), high-intensity functional training (22), strongman (13), and weightlifting (21), and its popularity among recreational athletes (4,27), evaluating the free-weight back squat is important. For recreational athletes, the joint motions and muscle actions involved in the back squat closely mimic those used in daily activities, suggesting that improvements in squat strength may enhance an athlete's quality of life (28). In contrast, competitive athletes require back squat adaptations that facilitate strength transfer to sport-specific movement patterns, which are often characterized by maximizing speed and power (33). Velocity-based training (VBT) systematically measures barbell kinematics during each repetition. Velocity-based training motivates athletes to apply maximal velocity consistently and guides accurate prescriptions of loads, sets, and repetitions, ultimately enhancing performance and competitiveness (40).

The barbell's speed and power in the concentric phase can be increased by enhancing the descent velocity, a factor examined in detail in 2 studies (2,36). Faster descent velocities caused significant increases in barbell kinematics (mean and peak velocities) compared with normal and slow descent at 60 and 80%, respectively, of the back squat 1-repetition maximum (1RM) (2). In the concentric phase, the ascent movement is usually not smooth; however, it includes a sticking region. The sticking region is defined with the first peak velocity (Vpeak1) and the minimal velocity afterward (Vmin) (38).

Vpeak1 appearance and magnitude are influenced by factors such as external load and lifting technique, with heavier loads and specific technical execution increasing the possibility of a visible Vpeak1. Vpeak1 may exceed Vpeak2 under certain conditions. Vpeaks correspond to distinct phases of the concentric movement and pre- and poststicking region, each involving different biomechanical loads and muscle activities (38). The back squat's concentric phase begins at the lowest position (turnaround point) and, depending on the barbell load and technique, progresses either directly to Vpeak2 or first to Vpeak1. The movement typically transitions through a sticking region after Vpeak1, reaching Vmin that marks the end of the sticking phase before progressing to Vpeak2 in a successful repetition (38). Figure 1 illustrates 2 typical velocity–time profiles, with and without a sticking region.

Figure 1.

The figure illustrates 2 representative velocity curves at 70% of 1RM, 1 with and 1 without the bounce technique. The eccentric and concentric phases are separated by the turning point (velocity = 0 m/s). For the bounce repetition, a velocity peak near the turning point (Vpeak1) is observed, followed by a decline (sticking region) that ends in the velocity minima (Vmin) and is absent in the no-bounce repetition. Both repetitions were performed with the same load. The velocity peak at the end of the concentric phase (Vpeak2) is visible for both conditions.

In addition, a study examined the barbell kinematics (Vmean, Vpeak1, Vpeak2, and Vmin), vertical ground reaction force (GRF), and muscle activity during a 4RM back squat test. The results revealed that Vpeak1 and Vmin were lower during slow descent, accompanied by a reduced peak GRF. There were no differences in Vpeak2 in the latter concentric phase. Furthermore, no variations in muscle preactivation were detected, suggesting that the diminished barbell kinematics may be attributed to factors including decreased potentiation, stored elastic energy utilization, and/or the stretch reflex (36). These mechanisms are categorized under the stretch-shortening cycle (SSC), in which active muscle lengthening is accompanied by active muscle shortening (1). Notably, several mechanisms contribute to the SSC effect, including neuromuscular preactivation, stretch–reflex responses, and elastic energy recoil stored in the tendons (31). During the early SSC's concentric phase, force, work, and power production were significantly enhanced compared with a purely concentric contraction without a preceding eccentric stretch (23,31,39).

Comparisons of guided back squat variations (including SSC, concentric-only, and concentric-only with isometric preload) have shown that SSC and isometric preload squats generate significantly higher work output within the first 300 milliseconds after the turning point than the concentric-only squats (39). This enhancement in early concentric performance highlights the potential benefits of incorporating SSC and isometric preload techniques to maximize the power output during the initial back squat phase. Regarding health implications, a goal-oriented descent velocity was recommended for prescribing, because there is a tradeoff between optimizing performance and decreasing injury risk (32). The inability to control the descent velocity of the eccentric phase, which might be comparable with an overly exaggerated bounce, can cause ballistic contact between the hamstring and thigh muscles, potentially increasing the shear force on the knee joint. Consequently, a controlled descent was recommended for recreational athletes to reduce the injury risk.

However, in free-weight back squats, athletes often use a bounce technique to overcome biomechanically disadvantageous positions while transitioning from the eccentric to the concentric phase. Therefore, rather than increasing velocity throughout the descent, accelerating only the eccentric phase final portion may provide a safer and more effective alternative. This strategy actively uses the SSC by increasing velocity toward the descent's bottom (39), thereby optimizing the transition into the concentric phase. The bounce technique uses active acceleration to enhance performance at this critical phase, which is contrary to traditional guidelines advocating controlled descent and avoidance of passive rebound at the lowest point (35).

To the best of our knowledge, no previous studies have specifically examined the bounce technique in free-weight back squats, differing from analyses of the full descent that do not characterize velocity near the turnaround point. Only a few studies have examined the bounce technique in other exercises. For bench press throws, a study investigated the training effects of the bounce technique during training and reported no significant differences in throwing velocity, 1RM, or power output (18). Another study assessed the bounce technique's barbell kinematics in the bench press throw and found significant improvement in the barbell kinematics (greater mean power [Pmean], Vmean, and time to peak power [Pmax]) (29). Furthermore, little is known about sex-specific effects during back squats. However, previous studies have reported differences in propulsive velocity (24), joint kinematics, and muscle activation (37). Therefore, recording movement characteristics, such as the bounce technique, might be beneficial, and appropriate methods must be developed. Wearable sensors enable real-time monitoring of physiologic and movement parameters, aiding injury prevention, load management, and training optimization (7). Recent advancements have improved the accuracy of detecting high-acceleration movements; however, further research is needed to determine the parameters that are most effective for enhancing performance and long-term athlete health (25). A potential parameter that describes the bounce behavior could be the GRF during the squat turning point.

There is a gap in the literature regarding the effects of the bounce technique on back squats, particularly regarding factors such as GRF and barbell kinematics. To our knowledge, no study has examined the influence of sex on the bounce technique. Therefore, we aimed to investigate the effects of the bounce technique on GRF and barbell kinematics during 2-legged full-back squats. We hypothesized that the bounce technique influences barbell kinematics positively and results in a higher GRF at the turning point. In addition, we aimed to assess the combined effects of descent velocity and bounce technique, while hypothesizing that the highest GRF and barbell kinematics would occur with this combination (36). Furthermore, we hypothesized that barbell kinematic measurements could predict the GRF value at the turning point.

Methods

Experimental Approach to the Problem

Subjects completed 2 experimental sessions using a within-subjects cross-sectional design to investigate the effects of bounce and descent velocities on GRF and barbell kinematics. In each session, the participants first determined their daily 1RM after a warm-up. Subsequently, 3 sets of exhaustion were performed (see Figure 1, Supplemental Digital Content 1, http://links.lww.com/JSCR/A694). In session 1, participants performed maximal-effort back squats using bouncing and nonbouncing techniques at 2 loads (70 and 80% 1RM). In session 2, participants completed back squats with fast and slow descent velocities using both techniques at 70% 1RM, as measured in session 1.

Subjects

Thirty-six participants (♀ = 16, ♂ = 20) completed session 1, and 29 participants (♀ = 13, ♂ = 16) completed session 2 of the bounce experiment (dropout rate, 19%). The withdrawal reasons were personal and unrelated to the study. Descriptive statistics showed that male participants had a mean age of 28.1 ± 4.2 years, body mass of 87.0 ± 7.6 kg, height of 178.7 ± 5.8 cm, 1RM of 164 ± 30 kg, and a 1RM relative to body mass of 1.9 ± 0.3. Female participants had a mean age of 28.0 ± 5.1 years, body mass of 71.1 ± 10.2 kg, height of 164.9 ± 5.6 cm, 1RM of 113 ± 19 kg, and a relative 1RM of 1.6 ± 0.2. We included participants who refrained from lower-body training for 72 hours before testing; had no injuries; did not use medications affecting metabolic or cardiovascular function; had ≥24 months of resistance training experience (minimum twice weekly), including back squats; and no history or current use of anabolic steroids. This study followed the ethical principles outlined in the Declaration of Helsinki for Human Experimentation and was approved by the regional ethics committee of Berne, Switzerland (2018-00742). Written informed consent was obtained from all participants before their participation.

Procedures

Subjects attended the laboratory twice for 4 weeks; each visit was ≥3 days. In session 1, we examined the effects of bouncing with 2 loads (70 and 80% 1RM). In contrast, in session 2, we investigated the effects of bouncing with 2 descent velocities (fast and slow) on the barbell kinematics and GRF.

The subjects completed a 10-minute individual warm-up before starting the experiment, including individual feedback on the VBT method. Subjects were instructed to apply the principles of VBT throughout the experiment, which emphasizes lifting with maximal intent: performing each repetition as fast as possible (9,40). Free-weight back squats were conducted on a Power Rack using an IWF-calibrated barbell and weight plates (Prestera Power Rack, XF Bar, and IWF Weightlifting Training Plates; Eleiko, Halmstad, Sweden). After the individual warm-up, a standardized warm-up was performed with loads 20, 40, 60, 80, and 90% of the estimated 1RM. In general, all loads were rounded to the next 2.5 kg. The 1RM test was completed by reaching a maximum of 5 tries, failing an attempt, compromised squat technique, or voluntary cancellation. Next, 3 sets to exhaustions that were not part of this study were executed. After the final set to exhaustion, the participants rested to ensure recovery. The physical demands of the technical and study phases were considerably less intense than those of the set-to-exhaustion portion. The details of the study protocol are shown in Figure 2 and Supplemental Digital Content 1 (see Figure, http://links.lww.com/JSCR/A694).

Figure 2.

Schematic representation of the study protocol for sessions 1 and 2. One- and 3-minute rest intervals were implemented between each repetition and each set, respectively.

In session 1, barbell kinematics and GRF for back squats with and without a bounce were determined across the 2 load conditions, beginning with the lighter load. The 2 load ranges represented a realistic weight range for an experienced strength-training population (19). The rest period given between repetitions was 1 minute, whereas 3 minutes was given between a change of technique or load conditions. In session 2, the bounce effect on 2 lowering velocities (fast and slow) was evaluated at a load corresponding to 70% of the 1RM. The participants self-selected the descent velocities, representing realistic training conditions. A repetition or technique was iterated if a repetition did not meet the standard criteria of VBT or if the technique deviated from an acceptable level to ensure accuracy and consistency in the collected data (5,9,11). An industry standard LPT (Seilzugsensor SX80-2500-1R-KA, WayCon, Munich, Germany) was used for the barbell kinematics and the velocity-based measurements, because it allowed for the exporting of raw data. The device was attached to the left side of the barbell between the rack and barbell sleeve. The LPT's (250 Hz) raw data were analyzed to calculate the barbell kinematics and its validity was ensured by comparison with the GymAware LPT in an unpublished material (Vmean: r = 0.99, root mean square error [RMSE] = 0.021 m/s; Vpeak: r = 0.099, RMSE = 0.026 m/s; Tconc: r = 0.99, RMSE = 0.062 s; range of motion: r = 0.99, RMSE = 0.007 m). Repetitions were segmented by identifying the negative displacement phase, including a local minimum that distinctly represented the squat movement's turning point. Subsequently, the local maxima on either side of these turning points were analyzed to determine the start and end of each repetition. Three-dimensional GRF data were directly measured using 2 force plates (Kistler 9260AA6; Kistler Instrumente AG, Winterthur, Switzerland), sampling at 1,000 Hz. Beginning each participant's measurement, both plates were calibrated and zeroed to ensure accuracy. Using cross-correlation, the signals between the LPT and force plates were synchronized and manually reviewed to ensure correctness.

Statistical Analyses

Statistical analysis was performed using the statistical computing environment R (26). With reference to the results of Saeterbakken et al. and α = 0.05 and β = 0.80, the sample size required to detect significant differences was 28 participants (29). Three trials of 1 repetition were conducted for each condition, and the trial with the median value was selected for the statistical analysis. The bounce effect on GRF and the 5 barbell kinematics (Pmean, Pmax, Vmean, Vpeak1, and Vpeak2) were evaluated using linear mixed effects (LME) modeling (p ≤ 0.05). Details on the independent variables are provided in Table 1. The first LME (LME1) accounted for the repeated measurements per participant with a random intercept. The models were built using these 3 variables (bounce [yes/no], sex [male/female], and load [70/80]) as fixed effects for session 1 and as a response cue (fast/slow) for session 2. In addition, an offset for the body mass was added for the 3 force-related independent variables (GRF, Pmean, and Pmax). The 3 force-related independent variables were best suited for multiplicative effects and were modeled using the log scale. In contrast, the 4 velocity-related independent variables (Vmean, Vpeak1, Vpeak2, and Vmin) demonstrated additive effects only (Supplemental digital content 2 for the model equations). Model assumptions and possible interactions were checked using residual analysis. Sessions 1 and 2 had similar magnitudes of the residual standard error and estimated variation for the random effects of the participants, demonstrating a reproducible study design. Another LME (LME2) was built to predict GRF using barbell kinematics. In LME2, the GRF estimated from the acceleration at the turning point served as an independent variable (see Text, Supplemental Digital Content 2 for the model equations, http://links.lww.com/JSCR/A693). The prediction's goodness was evaluated for the out-of-sample root using RMSE. The out-of-sample prediction was achieved through leave-one-out cross-validation and disregarding the random effects. Custom-written R script and associated data set are available on the Open Science Framework repository (URL: https://osf.io/fd7gv/).

Table 1.

Variables of interest.^*†

Variable	Description	Session 1 Mean ± SD	Session 2 Mean ± SD
GRF	Resulting ground reaction force at turning point (= lowest point of a back squat)	2,649 ± 652	2,456 ± 552
GRF/BW	GRF normalized by bodyweight (BW)	3.32 ± 0.53	3.14 ± 0.53
GRF/mass	GRF normalized by BW and external load	1.66 ± 0.31	1.66 ± 0.31
Pmean	Mean power measured during a (concentric phase) back squat with LPT	972 ± 265	1,007 ± 239
Pmax	Maximal power measured during a (concentric phase) back squat with LPT	2,430 ± 593	2,449 ± 609
Vmean	Mean velocity of the concentric phase of a back squat	0.53 ± 0.10	0.59 ± 0.08
Vpeak1	First velocity peak during the concentric phase of a back squat	0.55 ± 0.17	0.58 ± 0.11
Vpeak2	Second velocity peak during the concentric phase of a back squat	1.10 ± 0.19	1.19 ± 0.19
Vmin	Velocity minima between Vpeak1 and Vpeak2	0.39 ± 0.13	0.46 ± 0.11

^*This table contains a description of all force- and velocity-related variables used for the modeling.

^†GRF = ground reaction force.

Results

Significantly higher GRF values were observed at the turning point when the bounce technique was used, with a multiplicative factor of 1.19 (1.17, 1.21) and 1.22 (1.20, 1.24) for sessions 1 and 2, respectively (both p < 0.001). The multiplicative increase can be observed in the slopes of Figure 5, which got steeper with higher GRF values. In addition, Pmean, Vmean, Vpeak1, and Vmin increased with the use of the bounce technique. Meanwhile, Pmax and Vpeak2 were reduced in both sessions. These findings indicate a clear bounce effect on barbell kinematics and force production (Table 2, Figure 3, Figure 4).

Figure 5.

Effect of the bounce condition (B: Bounce, NB: No-bounce) on GRF estimated by the LME model (see Equation 1, Supplemental Digital Content 2, http://links.lww.com/JSCR/A693). The bold black line represents the mean effect, while the gray lines illustrate individual participant data. The slope of the connection lines visualizes the multiplicative effect found in the LME model.

Table 2.

Effects of bounce with 95% confidence interval and p-values.

Measurement	Session 1		Session 2
	Coefficient	p	Coefficient	p
GRF	1.19 (1.17 to 1.21)	<0.001	1.22 (1.20 to 1.24)	<0.001
Pmean	1.12 (1.05 to 1.10)	<0.001	1.12 (1.09 to 1.14)	<0.001
Pmax	0.91 (0.88 to 0.95)	<0.001	0.95 (0.92 to 0.99)	0.006
Vmean	0.04 (0.02 to 0.05)	<0.001	0.06 (0.05 to 0.08)	<0.001
Vpeak1	0.13 (0.11 to 0.15)	<0.001	0.14 (0.11 to 0.16)	<0.001
Vpeak2	−0.07 (−0.10 to −0.05)	<0.001	−0.05 (−0.08 to −0.02)	0.001
Vmin	0.03 (0.01 to 0.06)	0.003	0.05 (0.03 to 0.07)	<0.001

Figure 3.

Comparison of the vertical velocity curves recorded with a linear position transducer between bounce and no-bounce conditions. The curves represent the mean, while the boxplots illustrate the differences in the peak values and mean velocity during the concentric phase. (A) and (B) represent the velocity curves of both bounce conditions with 70 and 80% of 1RM, respectively. (C) and (D) represent the velocity curves of both bounce conditions with fast and slow decent conditions, respectively. The boxplot highlights differences of both conditions in barbell kinematics (m/s) during the concentric phase for each parameter. An asterisk (*) indicates a significance level of p < 0.001, extracted from the LME models.

Figure 4.

Comparison of inertial GRF at the turning point of the back squat normalized to the bodyweight of the participants. This facilitates cross-participant comparisons by accounting for individual differences in body mass and external load and solely visualizes the inertial part of the GRF.

Increasing the load from 70 to 80% 1RM resulted in a small but significant increase in GRF (coefficient = 1.03 [1.01–1.05], p < 0.001). All barbell kinematic variables decreased with higher loads: Pmean (0.84 [0.82–0.86], p < 0.001), Ppeak (0.96 [0.93–0.99], p = 0.02), Vmean (−0.13 [–0.12 to −0.14], p < 0.001), Vpeak1 (−0.10 [–0.08 to −0.12], p < 0.001), Vpeak2 (−0.11 [–0.08 to −0.14], p < 0.001), and Vmin (−0.17 [–0.15 to −0.19], p < 0.001).

Ground reaction force was not significantly affected when switching from a fast to a slow descent velocity occurred (coefficient = 0.99 [0.97–1.01], p = 0.45); however, it resulted in reduced Pmean (0.95 [0.93–0.98], p < 0.001) and Ppeak (0.93 [0.90–0.97], p < 0.001). Furthermore, Vmean decreased (−0.03 [–0.04 to −0.01], p < 0.001), and no significant changes were observed for Vpeak1 (−0.01 [–0.03 to 0.01], p = 0.465). Vpeak2 decreased significantly (−0.06 [–0.09 to −0.03], p < 0.001), whereas Vmin decreased slightly (−0.02 [–0.04 to 0.00], p = 0.02).

A significant sex effect was observed for Pmean in both sessions, with lower values in female participants (session 1: coefficient = 0.85 [0.78–0.93], p < 0.001; session 2: 0.89 [0.80–0.98], p = 0.019). In addition, a small effect was found for Vmean in session 1 (−0.04 [–0.08 to 0.00], p = 0.038). No significant sex differences were observed for GRF, peak power, Vpeak1, Vpeak2, or Vmin in either session (see Supplemental Digital Content 3, http://links.lww.com/JSCR/A693). Finally, we investigated whether GRF could be predicted from barbell kinematics. An absolute error of RMSE 101N was found when GRF was calculated using the acceleration at the turning point. A statistically significant systematic underestimation was observed in 71% of the cases. Calibration with LME2 improved RMSE 89N, demonstrating a slight enhancement in GRF prediction using barbell kinematic data.

Discussion

In this study, we primarily aimed to examine the bounce technique's impact on barbell kinematics and GRF. Hypothesis 1 proposed that the bounce technique would increase barbell kinematics and result in a higher GRF. Our findings showed that the bounce condition produced significantly greater velocity during the early concentric phase, as represented by Vpeak1, consistent with results from other bounce-related studies (23,36,39) and the SSC concept (6,12,35,41). This outcome confirmed that the bounce technique can effectively enhance barbell kinematics at the start of the lift. Notably, the bounce condition caused a brief reduction in Vpeak2 (Table 2 and Figure 2), which remains unexplained. However, this was only significant for session 1. One probable explanation is that the bounce technique generates a velocity drop or “sticking region,” because of a disproportionate effect on velocity parameters: Vpeak1 increases more than Vmin when bouncing (0.13 vs. 0.03 m/s). In addition, Vpeak1 decreases less than Vmin (−0.10 m/s vs. −0.17 m/s) with increasing load, thereby amplifying the velocity drop (Figure 2A, B). This observation of Vpeak1 and Vpeak2 among the bounce conditions highlights the need for clear variable differentiation. In addition, Pmax was reduced when the bounce was used, which possibly showed the correlation between power and peak velocity during the later phase of the lift. Collectively, these insights emphasized that programming a bouncing technique to achieve the desired training adaptations may depend on the load, individual techniques, and anthropometrics. Similarly, the bounce condition resulted in significant increase in Vmean and Pmean (29). Expectedly, in contrast to the effect of the bounce technique, increased load showed a consistent impact on all velocity parameters (29), with the largest effect for Vmin. These findings underscore the importance of not overlooking the bounce technique when applying VBT because differences in barbell kinematics may lead to different training prescriptions.

In both training sessions, back squats performed with the bounce technique significantly increased the GRF compared with squats performed without the bounce. No interaction was observed between the load and bounce technique, which contrasted with the initial expectation that higher loads may enhance the bounce effect at the turning point. Moreover, it was anticipated that heavier loads might prompt technique alterations among the participants; however, this was not observed. A slight increase in the GRF was noted with heavier loads (1.03 [1.01–1.05]); however, this was below the anticipated level and was less impactful than the bounce effect. This aligned with the observation that sex had no significant effect on barbell kinematics or GRF. This finding suggested that the load deceleration at the turning point contributed substantially to the GRF than the static force of the load.

Next, a large interindividual variability in GRF was found, which can be observed in Figure 5. For some participants, these differences exceeded ×1 body weight, which was unexpectedly large, highlighting the need for goal-oriented programming of the bounce technique. Considering that these elevated forces may pose additional health risks (32), practitioners should carefully weigh the benefits of exploiting the bounce effect against the potential downsides. Increasing the external load (from 70 to 80% 1RM) seems to have a limited direct effect on the GRF. However, monitoring the technique is crucial. Unintended technique shifts caused by the load could introduce a bounce pattern and significantly increase the GRF. Regarding barbell kinematics, practitioners should consider the relationship between the load and barbell kinematics to determine whether the bounce technique achieves the intended outcome and precisely identify which portion of the concentric phase should be modulated. Furthermore, the risk–reward ratio of the GRF in relation to the stress imposed on the body should be assessed. The effects of the bounce technique and load are highly individualized; therefore, no definitive recommendations can be made without further research. These findings highlighted the need for cost-effective in-field measurement devices to confirm whether the intended stimulus was achieved through technical modifications.

Contrary to our expectations in hypothesis 2, no significant effect of descent velocity on GRF was detected, nor was there any significant interaction between descent velocity (slow or fast) and the bounce condition (Figure 3). This finding was unexpected, especially because previous research (2,36) demonstrated that a faster descent increases barbell kinematics, suggesting that it might have a similar effect on the GRF. Unlike these studies, where fast descent speeds likely triggered the SSC reflex (bounce technique), we controlled the bounce at the turning point for both slow and fast descents. For descent (fast, slow), no significant effect was observed in the early concentric phase (Vpeak1); however, a faster descent resulted in a small increase in the latter part of the concentric phase (Vpeak2) and, consequently, in Vmean. This increase in the barbell kinematics in the latter part of the concentric phase could be driven by several factors. One possibility is that a faster descent makes fully neutralizing the bounce difficult, especially when paired with a deliberate bounce condition. However, no interaction was observed between the bounce condition and descent velocity. Another explanation involves technique adjustments; participants may adopt a more rigid torso and steadier hip/knee angles during rapid descent, potentially improving concentric efficiency. Ultimately, the precise mechanisms underlying increased barbell kinematics in the latter part of the concentric phase remain unclear and necessitate further investigation. Our findings indicate that different descending velocities without the bounce technique do not affect the GRF or early barbell kinematics (Vpeak1, Vmin). However, they have a small effect on the latter part of the concentric phase (Vpeak2), which practitioners should consider when required. We did not observe any significant effect of sex on barbell kinematics or GRF. This was an unexpected result for the GRF as female participants lifted lighter loads; therefore, a difference could have been allegeable. However, this indicates that the deceleration of the load at the turning point contributes more substantially to the GRF than the static force of the load. A lower mean power was recorded for female participants, which may be attributed to the lighter loads lifted.

We hypothesized that barbell kinematic measurements could predict the GRF value at the turning point. Consistent with previous findings (16), a statistically significant systematic underestimation was observed compared with the measured GRF data. This is potentially sufficient for monitoring squat mechanics and load management; however, further research is required to validate its use across different contexts. Health-focused practitioners can use it to monitor stress during squats, whereas performance-focused practitioners can track consistency and techniques. Therefore, this approach enables VBT monitoring tools to predict vertical GRF at a key point in the back squat.

A limitation of this study is that all participants performed the squat in a prefatigued state, thereby influencing the squat technique and barbell kinematics, particularly in the sticking region. In addition, the fixed order of bounce and no-bounce conditions for all participants may have introduced a potential familiarization effect. However, the residual standard error and estimated variation suggested no learning transfer from sessions 1 to 2, demonstrating a reproducible study design. Second, no fatigue effect was found in the sets between repetitions 1–3 for both sessions and all conditions (see https://osf.io/fd7gv/). In addition, a significant effect of descent velocity was found in session 2, with the fast condition showing a significant increase in Vmean. Notably, the fast condition was tested later than the slow condition. Assuming some fatigue accumulation over time, the observed difference may be conservative without fatigue, or the effect could be larger. In this study, we used a narrow load range, which limited the scope of the findings. Expanding the load spectrum in future studies can provide more comprehensive insights. Conclusions on barbell kinematics and GRF were obtained; however, studying the effect of the bounce condition on the body in the form of muscle activity, muscle force, or joint moments was not possible. Future studies should include a larger loading spectrum, 3-dimensional motion-capture measurements, and inverse kinetics to establish more detailed information on the effects of bounce conditions on the body.

Practical Applications

These findings indicate that using the bounce technique significantly increases the GRF at the turning point and is accompanied by enhanced early concentric barbell velocity (Vpeak1). The bounce can help overcome a sticking region early in the lift. However, its associated increase in the GRF underscores the importance of balancing performance gains with load management considerations. No sex differences were observed in GRF or barbell kinematics. Furthermore, increasing the load from 70 to 80% of 1RM introduces a minor additional effect on the GRF, provided the technique remains consistent. This finding suggests that practitioners can safely increase loads within this range without substantially increasing the physical demands on the athlete. Regarding barbell kinematics, the bounce technique increased early concentric velocity (Vpeak1) more than Vmin, causing a significant velocity drop in the sticking region. Similarly, the effect of load decreased Vmin more than the other velocity parameters. Furthermore, Vpeak2 decreased markedly with the bounce technique but increased with a faster descent velocity. This observation indicates that while the bounce technique enhances velocity during the initial lift phase, it can reduce it in the later phase. Therefore, adopting a faster descent velocity and reducing the load can offset the decrease in the late-phase velocity, thereby maximizing the power output. These contrasting effects highlight the need for practitioners to strategically apply techniques and loads in a goal-oriented manner to optimize outcomes. In addition, predicting the GRF using barbell kinematics could provide valuable insights for the use of VBT monitoring tools and offer practical applications for health and strength practitioners, facilitating improved monitoring of techniques, consistency, and physical stress during training.

Supplementary Material

jscr-40-001-s001.docx ^{(5.2MB, docx)}