Does Side Dominance Affect the Symmetry of Barbell End Kinematics during Lower-Body Resistance Exercise?

Abstract

Lake, JP, Lauder, MA, and Smith, NA. Does side dominance affect the symmetry of barbell end kinematics during lower-body resistance exercise? J Strength Cond Res 25(3): 872-878, 2011-The aim of this study was to examine whether ground kinetic asymmetries significantly affected measures of bar end power output during bilateral lower-body resistance exercise. Vertical ground reaction forces (GRFs) from both feet and power outputs from both ends of the bar were recorded using force platforms and high-speed cameras simultaneously during the back squat performance of 10 experienced male volunteers with 30, 60, and 90% of their 1 repetition maximum. Side dominance was determined according to perceived left- and right-side handedness, dominant left- and right-side GRFs, and barbell power output side dominance. The dominant and nondominant side average concentric phase GRFs and barbell end power outputs were compared at each loading condition using 2-way analysis of variance. Pearson product-moment correlations were also performed to assess the relationship between average GRF and bar end differences. The results showed that although differences between the dominant and nondominant side GRFs reached 21% in some cases, there were no significant differences between the left- and right-side bar end power outputs, which did not exceed 3.4%. There were no consistent correlations between the dominant and nondominant side average GRFs and average bar end power differences. In addition to this, progressive loading did not significantly affect differences between the dominant and nondominant side GRFs or bar end power output differences. It was concluded that although apparently healthy individuals demonstrate considerable side dominance at the foot-floor interface, these tend not to be reflected in the symmetry of bar end power output.

Introduction

There has been a recent increase in the research focus on movement symmetry during bilateral resistance exercise (^5,18,21). The study of independently measured left- and right-side ground reaction forces (GRFs) has shown that during controlled bilateral resistance exercise, healthy individuals tend to favor a side that may not correspond with the side they perceive to be dominant by as much as 10% (^5,18). Some researchers have suggested that this may be an underlying cause of injury (⁵), whereas others have suggested that it may be a consequence of past injury or leg length discrepancy (¹⁸).

Lauder and Lake (¹³) recently demonstrated that during power snatch performance, asymmetric intervention significantly influenced bar end trajectory. However, little is known about whether side dominance, determined from independently measured GRF, influences the symmetry of left- and right-bar end kinematics.

This is an important but apparently overlooked aspect of measurement methodology that could have important implications for strength and conditioning practitioners. The accurate measurement of resistance exercise mechanical power output is critical for monitoring both resistance training intensity (^{1,2,12,15,16,23}) and the effects of resistance training (^4,10,11), and to varying extents relies on bar end kinematics (^{2,3,6,7,9,14,17}). Therefore, a side dominance effect could significantly influence measures of resistance exercise mechanical power output that are obtained from bar end kinematics. This could in turn influence training load prescription and performance monitoring.

The aim of this study therefore was to test the hypothesis that ground kinetic asymmetries would significantly affect the symmetry of bar end power output. A secondary aim of this study was to test the hypothesis that progressive loading would intensify this effect.

Methods

Experimental Approach to the Problem

Ten male volunteers with 1 year's back squat experience participated in 2 testing sessions that were separated by approximately 7 days: session 1, during which the back squat 1 repetition maximum (1RM) was established, and session 2, during which asymmetry testing was performed. During asymmetry testing, subjects performed 2 maximal effort back squats with 30, 60, and 90% of their 1RM to encompass a light, medium, and heavy spectrum of relative exercise intensity. The GRFs of both feet were recorded independently using 2 Kistler force platforms, whereas the movement footage of both bar ends was recorded from 3 high-speed digital cameras. The left- and right-side GRFs and bar end power outputs were averaged across the concentric phase of back squat performance, and side dominance was determined using 3 different methods: perceived handedness (left- or right-side dominance: LRSD), left and right GRF side dominance (force side dominance [FSD]), and left and right bar end power output side dominance (barbell side dominance [BSD]), where side dominance was defined as the highest left- or right-side average GRF or bar end power. The effect that the independent variables (side dominance and progressive loading) had on the dependent performance variables (average GRF and average bar end power) was then analyzed using 2-way analysis of variance. In addition to this, Pearson product-moment correlations were performed to test the relationship between AGRF and ABP dominant and nondominant differences.

Subjects

Ten physically active men with a minimum of 1 year's back squat experience volunteered to participate in this study. Their mean (± SD) physical characteristics were age: 28.8 (±8.5) years, mass: 80.6 (±10.7) kg, height: 180 (±0.04) cm, squat 1RM: 122.3 (±36.7) kg, and relative squat 1RM: 1.5 (±0.4) times body mass.

University of Chichester ethics approval was obtained before data collection, and all participants completed a health history questionnaire and provided written informed consent.

Procedures

All subjects participated in 2 testing sessions that were separated by approximately 7 days: the first session, during which the back squat 1RM was established using a procedure that was similar to that outlined and used by Stone et al. (²²), and a second session, during which asymmetry testing was performed.

During both testing sessions, the measurement of back squat performance began after a loaded barbell (Eleiko Weightlifting Training Bar, Sweden) positioned across the subject's posterior deltoids immediately below the C7 vertebrae (^9,22) was taken from free-standing squat stands (Scorpion Gym Equipment, Nottingham, United Kingdom). The participant squatted until the upper surface of the thigh was parallel with the ground and stood upright to the start position to complete the lift (^20,22). The parallel position depth was gauged visually during the 1RM testing session and the bottom position recorded. Using this information, a bungee cord was positioned across a free-standing frame to enforce consistent depth (²⁰). Any squats that did not meet the depth criteria were excluded from the analysis. Participants were instructed to perform the eccentric phase of the back squat under control and perform the concentric phase as explosively as possible while maintaining foot contact with the ground.

During asymmetry testing, each participant performed 2 maximal effort single back squats with 30, 60, and 90% 1RM with a minimum of 1 minute and a maximum of 3-minute recovery between each lift (¹⁹). The 30, 60, and 90% 1RM loads were selected to encompass a light, moderate, and heavy spectrum of relative exercise intensity. Verbal encouragement was given during all performances (¹⁰).

Measurements

The vertical GRF of back squat performance was recorded from both feet individually by 2 0.4 × 0.6-m Kistler 9851 force platforms (Alton, United Kingdom) at a sampling frequency of 500 Hz. The analog GRF signals were amplified by 2 type 9865E 8-channel charge amplifiers before they were digitally converted.

Three digital cameras (Basler A602fc-2, Ahrensburg, Germany) were positioned on rigid tripods around and approximately 5 m from the center of the area of interest (Figure 1). Each camera filmed back squat performance at 100 Hz with a shutter speed of 1/1,000 seconds (⁸) after first recording a 17-point calibration frame (Peak Performance Technologies Inc., Englewood, CO, USA). The GRF and bar end kinematics were synchronized using a Vicon MX control unit (Peak Performance Technologies Inc.).

Figure 1: A schematic of the experimental setup that shows the position of the 3 cameras and 2 force platforms relative to the position of the bar during back squat performance.

Retroreflective markers that were positioned on both ends of the bar were digitized at 100 Hz using Peak Motus 9.2 software from approximately 10 frames before the conclusion of the eccentric phase to approximately 10 frames after the conclusion of the concentric phase. After digitization, the raw coordinate data were smoothed using a low-pass filter with a cut-off frequency of 6 Hz (⁸) and differentiated with respect to time to obtain bar end velocity and acceleration.

Bar end power was calculated by multiplying bar end force by its velocity. Bar end force was calculated using the methods recently described and used by Hori et al. (⁹) where the mass of the bar was multiplied by its acceleration and added to the product of bar mass multiplied by gravitational acceleration (g):

Measures of GRF and bar end power were then averaged across the duration of the concentric phase for further analysis. This approach has recently been used by Flanagan and Salem (⁵), who suggested that peak performance data may not accurately represent the behavior of parameters of interest over a selected period of time. The concentric phase was determined from the bar end velocity-time curve, with the transition from negative to positive velocity indicating the beginning of the concentric phase and the transition from positive to negative velocity the end of the concentric phase (⁵).

Side dominance was determined using 3 different methods: perceived handedness (left-right side dominance: LRSD) (^5,18); left- and right-side concentric phase GRF dominance (FSD) (^5,18); and left and right concentric phase bar end power output dominance (BSD). Differences between the left and right and dominant (D) and nondominant (ND) side average GRF (AGRF) and average bar end power outputs (ABP) were then calculated as percentage differences using standard procedures.

Statistical Analyses

The absolute and relative measurement reliability of the AGRF and ABP was assessed using the coefficient of variation (CV) and intraclass correlation coefficients (ICCs) respectively on within session test-retest data obtained from back squat performances with 30, 60, and 90% 1RM. To test the hypotheses that ground kinetic asymmetries would significantly influence bar end symmetry and that progressive loading would intensify this effect a 2-way (side × load) analysis of variance was used to examine mean differences in the AGRF and ABP. In addition to this, Pearson product-moment correlations between the D and ND side differences for each of the different methods and loads were calculated to provide a descriptive view of the relationships between ground kinetic asymmetries and bar end power symmetry. All statistical calculations were performed using SPSS version 16.0 for Windows (SPSS, Inc., Chicago, IL, USA), and an alpha value of p ≤ 0.05 was used to determine statistical significance.

Results

The results of the test-retest analysis are presented in Table 1 and demonstrate a high degree of both relative and absolute reliability for AGRF and ABP at different relative intensities.

Table 1: Mean % differences, CV and ICC for the measures of AGRF and ABP at 30, 60, and 90% 1RM.*

The mean (± SD) D and ND side concentric phase AGRF and ABP are presented in Table 2, and the mean percentage differences between the D and ND side AGRF and ABP are presented in Table 3. Differences between the D and ND side FSD AGRF were statistically significant (F(1,54) = 8.39, p = 0.005, η² = 0.14, 1 − β = 0.81) However, this effect was not reflected in the differences between the D and ND FSD side ABP (F(1,54) = 0.018, p = 0.018, η² = 0.001, 1-β = 0.894) and was not influenced by progressive loading (F (2,54) = 0.055, p = 0.946; η² = 0.002, 1 − β = 0.05). The remaining AGRF and ABP D and ND side differences did not reach statistical significance (p > 0.05), and ranged between −6.2 to 9.3% for the AGRF and −1.4 to 3.4% for the ABP across the different relative intensities (Table 3).

Table 2: Mean (± SD) D and ND side concentric phase AGRF and ABP.*

Table 3: Mean (± CI) percentage differences between the D and ND side concentric phase AGRF and ABP.*

The relationships between D and ND AGRF and ABP differences are presented in Table 4 and in Figures 2-4. At 30% 1RM, there was a strong but nonsignificant negative relationship (FSD: r = −0.63, p > 0.05; LRSD: r = −0.59, p > 0.05; BSD: r = −0.60, p > 0.05) between the AGRF and ABP D and ND side differences, with increases in these differences resulting in no change or a reduction in the ABP D and ND side differences (see Figures 2-4). The relationship between the AGRF and ABP D and ND side differences were negligible for all methods at 60% 1RM and for FSD and BSD differences at 90% 1RM (see Table 4 and Figures 2-4). However, at 90% 1RM the LRSD D and ND side AGRF and ABP differences were significantly related (r = 0.66, r² = 0.43, p < 0.05) (Figure 3).

Table 4: The relationship between the AGRF and ABP D and ND side differences.*

Figure 2: The relationship between the force side dominance average ground reaction force and average bar power dominant and nondominant side differences at 30, 60, and 90% 1RM.

Figure 3: The relationship between the left-right side dominance average ground reaction force and average bar power dominant and nondominant side differences at 30, 60, and 90% 1RM.

Figure 4: The relationship between the bar power side dominance average ground reaction force and average bar power dominant and nondominant side differences at 30, 60, and 90% 1RM.

Discussion

The results of this study demonstrated that asymmetries in ground kinetics did not influence the symmetry of the bar end power output, leading to the rejection of the first hypothesis. Statistically significant differences of between 13.5 and 20.7% were found between the D and ND side AGRFs when side dominance was determined according to the dominant left and right AGRFs (FSD); differences that were considerably greater than those previously reported (∼6%: 5,18). The results indicated that the increased technical demands of heavier back squat performance (60 and 90% 1RM) reduced the relative D and ND side AGRF differences. Although not statistically significant, the relative consistency of the loading effect observed during the 60 and 90% 1RM conditions suggested that the assessment of ground kinetic asymmetry during bilateral resistance exercise must consider the potential effects of progressive loading.

Interestingly, when side dominance was determined according to perceived handedness (LRSD), differences between the D and ND AGRFs were consistent with the findings of Newton et al. (¹⁸) in terms of both magnitude and a lack of statistical significance. However, the differences observed in this study were not consistent across the different loading conditions (30%: 5.6%; 60%: −2.5%; 90%: 8.2%), suggesting that although the load affect was not statistically significant, perceived handedness may not be the most reliable way to determine side dominance for the assessment of movement symmetry during bilateral lower-body resistance exercise. Of course, the way in which side dominance is determined will depend largely on the facilities available, but it appears that the effective assessment of ground kinetic asymmetries requires the ability to independently measure left- and ride-side GRF. The strong positive relationship that was found between the D and ND LRSD AGRF and ABP differences at 90% 1RM was interesting but of little practical relevance in an applied perspective when the inconsistent nature of the other relationships was considered. However, this may have been a consequence of the large variability that was observed in this study. Further study into this aspect of the study using a single-subject design may be justified and may provide an insight into the predictive ability of bar end asymmetries on ground kinetic asymmetries.

The differences that were observed between the dominant left- and right-side AGRFs (FSD) and the side that was perceived to be dominant (LRSD) did not influence the symmetry of bar end power outputs (FSD: −0.1 to 2.2%; LRSD: −1.4 to 1.6%). In fact, the greatest mean difference that was observed between the left- and right-side ABP (BSD) during this study was 3.4% (1.7-5.1% 95% confidence interval). From an applied perspective, such a difference is not a concern but is a surprise given the large ground kinetic asymmetries that were observed.

A graphical example of good ground kinetic and good bar end symmetry is presented in Figure 5, whereas an example of poor ground kinetic symmetry and good bar end symmetry is presented in Figure 6. Figure 5 illustrates a concentric phase ground kinetic asymmetry that did not exceed 2% and bar end kinematic asymmetry that remained under 0.6%, although bar end asymmetries did reach 8.4% in some subjects with mode values of ∼5%. Figure 6 illustrates a concentric phase ground kinetic asymmetry that averaged ∼7% but reached 18% at its peak. This difference was typical of the FSD GRF differences. The pattern of bar end power output differences was also typical, with differences not exceeding 0.6%.

Figure 5: An example of good ground reaction force and bar end kinematic symmetry during back squat performance.

Figure 6: An example of poor ground reaction force symmetry and its lack of effect on bar end kinematic symmetry during back squat performance.

These findings are unique to this study because the effect that ground kinetic side dominance has on the symmetry of bar end power outputs had not previously been investigated. They are important because (a) they demonstrate that ground kinetic asymmetries do not affect the symmetry of bar end power output, which in turn supports the efficacy of methods that rely on the kinematics from one bar end to obtain measures of resistance exercise mechanical power output and (b) they indicate that the body must compensate in some way to avoid the quite considerable ground kinetic asymmetries effecting the symmetry of the barbell, which may go some way to support the contentions of Newton et al. (¹⁸).

To conclude, ground kinetic asymmetries and progressive loading do not significantly affect the symmetry of bar end power output supporting the efficacy of methods that rely on bar end kinematics to obtain resistance exercise mechanical power output (^{1-3,7,9,14,15,20,22,23}). They have important implications for strength and conditioning professionals because this method tends to be more accessible and less technically complex and financially costly compared with alternative and typically laboratory-based methods.

Practical Applications

Apparently healthy individuals demonstrate considerable differences between the GRFs that are generated between the left and right sides during back squat performance. These may be a cause for concern and should be treated accordingly. However, they do not affect the symmetry of the barbell and as such the measures of resistance exercise power that are often obtained from bar end kinematics. In addition to this, progressive loading does not significantly influence ground kinetic or bar end kinematic side differences. However, if ground kinetic side differences are of interest, it is suggested that loads that are greater than 30% 1RM are used. In addition to this, ground kinetic side differences should be assessed from independently measured left- and right-side ground reaction forces and the symmetry of bar end power from independently measured left and right bar end kinematics.