Velocity-Based Resistance Training Monitoring: Influence of Lifting Straps, Reference Repetitions, and Variable Selection in Resistance-Trained Men

Abstract

Background:

Using lifting straps during pulling exercises (such as deadlift) may increase absolute velocity performance. However, it remains unclear whether lifting straps could also reduce the degree of relative fatigue measured by velocity decline and maintenance in a training set.

Hypothesis:

There will be less mean velocity decline (MVD) and greater mean velocity maintenance (MVM) for deadlifts performed with (DLw) compared with without (DLn) lifting straps, and an underestimation of MVD and MVM when using the first compared with the fastest repetition as a reference repetition.

Study Design:

Randomized cross over design.

Level of Evidence:

Level 3.

Methods:

A total of 16 resistance-trained men performed a familiarization session, 2 1-repetition maximum [1RM] sessions (1 with and 1 without lifting straps), and 3 randomly applied experimental sessions consisting of 4 sets of 4 repetitions: (1) DLw against the 80% of DLn 1RM (DLwn), (2) DLn against the 80% of the DLn 1RM (DLnn), and (3) DLw against the 80% of the DLw 1RM (DLww). MVD and MVM were calculated using the first and the fastest repetition as the reference repetition.

Results:

MVD was significantly lower during DLwn and DLnn compared with DLww (P < 0.01), whereas MVM was greater during DLwn and DLnn compared with DLwn (P < 0.01) with no differences between DLwn and DLnn for both MVD and MVM (P > 0.05). The second repetition of the set was generally the fastest (54.1%) and lower MVD and higher MVM were observed when the first repetition was used as the reference repetition (P < 0.05).

Conclusion:

Lifting straps were not effective at reducing MVD and increasing MVM when the same absolute loads were lifted. Furthermore, using the first repetition as the reference repetition underestimated MVD, and overestimated MVM.

Clinical relevance:

The fastest repetition should be used as the reference repetition to avoid inducing excessive fatigue when the first repetition is not the fastest.

Resistance training (RT) is arguably the most effective mode of exercise for improving muscular strength, power, and hypertrophy, with these adaptations often coinciding with improvements in jumping, sprinting, agility, and sport-specific tasks.^26,27 During RT, coaches often instruct athletes to perform concentric muscle actions as fast as possible since explosive concentric muscle actions often result in greater training adaptations compared with intentionally slower concentric actions.^8,21 In that regard, many different strategies have been used to ensure high-velocity outputs during RT such as manipulating set structures (ie, using cluster or rest redistribution set structures),^11,17 providing velocity feedback,^22,23 and using ergogenic aids such as caffeine. ²⁴

Interestingly, Jukic et al ¹⁰ recently showed that even the simple use of lifting straps during deadlifts could provide ergogenic effects on absolute velocity outputs. Although this sounds desirable from a performance-enhancement perspective, the amount of fatigue experienced (eg, velocity decline) was not examined in that study. Lifting straps have been shown effective to increase absolute performance (eg, 1-repetition maximum [1RM] strength or lifting velocity) during exercises where grip strength may be a limiting factor.^3,10,13 It remains unclear whether lifting straps could also reduce the degree of relative fatigue experienced in a training set when the performance-limiting aspect of the exercise (eg, grip strength) is removed. Objectively assessing velocity during cases like this becomes extremely relevant, especially if velocity-based training practices are adopted.

Although the idea of using velocity to assess performance and adjust training prescriptions is no longer novel, the way in which velocity is used to prescribe RT is quite inconsistent. In some cases, strength and conditioning practitioners may choose to use the first repetition of a set as the baseline measure to which all other repetitions are compared, but others may choose to use the fastest repetition, regardless of where it occurs during the set.^6,7,16 Although studies have used each of these independently, there is a lack of research that directly compares how each of these can affect real-life training decisions if velocity-loss thresholds are used. Furthermore, velocity-loss is often described as a decrease in performance (ie, velocity decline), which can be quite extreme in some cases when performance declines quickly near the end of an exercise set. However, when using the velocity decline to describe how fatiguing an exercise was, the large majority of repetitions are ignored because only the fastest (or first) and last repetitions are considered, without taking into account the overall velocity maintenance of all of the repetitions performed. Therefore, overall velocity maintenance might also serve as a metric worth monitoring as it provides a more holistic view of fatigue within a set.

To show how these decline and maintenance variables differ, 1 previous study using moderate-intensity, high-volume back squats showed that for the same training stimulus, velocity decline (comparing the first to the last repetition) was over 23%, whereas the overall velocity maintenance (the average of all repetitions compared with the first) was 92%. ³⁰ Therefore, failing to consider velocity maintenance, and calculating only velocity decline, might overestimate the degree of fatigue induced by RT, which is important to consider when using velocity as a measure of fatigue. Although some evidence exists to support this notion, research on heavier loads with less training volume in different exercises has yet to be conducted.

This distinction between velocity decline and maintenance becomes more important as recent evidence suggests that analyzing the absolute mean values of a given mechanical variable is not the same as assessing the change in that particular variable throughout RT sessions. ¹¹ In that regard, the effects of external factors like lifting straps on velocity outputs might vary as a function of the variable being analyzed (ie, absolute vs decline vs maintenance). This becomes even more relevant since, although velocity decline is frequently used to both regulate RT volume and to quantify the acute effects of various RT protocols, the reference repetition used to calculate velocity decline is not consistent across studies (or in practice) and external factors like lifting straps may complicate matters even more. In that regard, a repetition-by-repetition analysis could potentially answer this question by indicating on how many instances a certain repetition in the set was the fastest. Furthermore, the effect of using the first or the fastest repetition for computing velocity maintenance and decline variables have not been compared within the same study.

To address gaps in the literature, we compared velocity decline and maintenance between deadlifts performed with (DLw) and without (DLn) lifting straps under 3 conditions: (1) DLw against the 80% of DLn 1RM (DLwn), (2) DLn against the 80% of the DLn 1RM (DLnn), and (3) DLw against the 80% of the DLw 1RM (DLww). In addition, we also investigated the number of instances during which the first repetition in the set was also the fastest across the deadlift conditions. Based on previous findings,^7,10,16 we hypothesized (1) greater velocity decline during DLnn and DLww than during DLwn, (2) greater velocity maintenance during DLwn than during DLnn and DLww, and (3) an underestimation of velocity decline and maintenance when using the first repetition as a reference repetition since the first repetition was not expected to be the fastest most of the time.

Methods

Study Design

A randomized cross over design was used to (1) examine the effects of lifting straps on velocity decline and maintenance during multiple sets of the deadlift exercise, (2) examine the effects of the reference repetition on the magnitude of velocity decline and maintenance, and (3) perform a repetition-by-repetition analysis to identify the number of instances when each of the repetitions in every set was the fastest. Three preliminary sessions included a familiarization session and 2 1RM assessment sessions (ie, 1 with and 1 without lifting straps). The 3 experimental sessions each included a different protocol that were performed in random order during 2 consecutive weeks, separated by 48 to 72 hours of rest. Each protocol included 4 sets of 4 deadlifts with 3 minutes of interset rest. In 2 sessions (DLwn and DLnn), subjects exercised with a load equal to 80% of their DLn 1RM, whereas in the remaining session (DLww), subjects exercised with a load 80% of their DLw 1RM. Subjects were instructed to refrain from any strenuous activity for at least 48 hours before each session and were allowed to use weightlifting chalk but lifting belts, knee wraps, the hook grip, and alternate grip were forbidden. The ecological validity of this approach is justified because the majority of our subjects reported to regularly use the double overhand grip with a conventional style in their training.

Subjects

A total of 16 men participated in this study (age, 24.4 ± 2.3 years; height, 181.6 ± 5.8 cm; body mass, 86.6 ± 8.2 kg). Subjects had at least 1 year of experience with heavy-load RT, were able to perform the DLn with at least 1.5 times their body mass (DLn 1RM, 162.4 ± 26.9 kg [1.88 ± 0.24 kg per kilogram of body mass]) and were not exclusively using lifting straps during training. All subjects reported to be free from any musculoskeletal injury that could compromise the testing procedures. Each subject received detailed information about the testing procedures and gave written informed consent before participating in the study (approval number 010/2019).

Familiarization Session

Body mass and height were assessed and then subjects completed a warm-up protocol that consisted of running on a treadmill for 5 minutes at 8 kmph, performing dynamic stretching for 5 minutes, and finally performing 2 sets (1 with and 1 without lifting straps) of 3 repetitions with 50% and 60% of their self-reported DLn 1RM. The position of the toes and heels as well as hands were self-selected and recorded with 2 measuring tapes placed on the floor in front and behind the barbell for each subject during all sessions to ensure identical starting stance each time. Finally, subjects estimated their deadlift 1RM with and without lifting straps and all of them estimated that 1RM would be greater with lifting straps. Based upon these estimations, the warm-up loads for each of the upcoming 1RM assessments were determined.

1RM Sessions

Subjects performed the same warm-up and then performed the 1RM protocol. The 1RM protocol consisted of 3 repetitions at 20%, 3 repetitions at 40%, 3 repetitions at 60%, 1 repetition at 80%, and 1 repetition at 90% of self-reported DLn 1RM followed by 1RM attempts.^12,14 After each successful 1RM attempt, the load was progressively increased until reaching the 1RM. Three minutes of interset rest were provided between attempts. Subjects were encouraged verbally to perform the concentric part of each repetition as explosively as possible and to lock out at the top as fast as possible.

Experimental Sessions

Subjects performed the DLn, DLw, and DLww sessions in random order, each separated by 48 to 72 hours of rest. Each experimental session was preceded by the same standardized warm-up described above. In addition, subjects performed 3 repetitions with 20%, 40%, and 60%, followed by 1 repetition with 80% of DLnn 1RM with 3 minutes of rest being provided between the sets. To ensure a consistent range of motion for all repetitions, subjects were required to avoid bouncing the loaded barbell off the floor when transitioning from 1 repetition to the next by implementing a 1 second pause with the barbell on the floor. Subjects were instructed to perform the lifting phase of all repetitions as fast as possible.

Data Acquisition and Analysis

A linear position transducer (LPT) (GymAware Power Tool, Kinetic Performance Technologies) was used to collect the mean velocity (MV) of all repetitions. This LPT, which has been shown to be valid and reliable,^5,9,28 was attached to the right side of the barbell between the hands and the loaded barbell sleeves. The LPT used in the present study measures the total displacement of its cable in response to changes in the barbell position and incorporates an angle sensor that accounts for motion in the horizontal direction during predominantly vertical displacement measurements. The LPT software later accounts for the total distance and angle, and, using basic trigonometry, provides a resultant vertical displacement. Instantaneous velocity was determined as the change in barbell position with respect to time, which is also provided by the LPT software. Data obtained from the LPT were transmitted via Bluetooth to a tablet (iPad, Apple Inc.) using the GymAware v2.4.1 app, and to the online cloud before being exported to Microsoft Excel (Microsoft Corporation) and prepared for further analysis. Using the MV data, MV decline (MVD) and MV maintenance (MVM) were calculated using the first and the fastest repetition of each set as the reference repetition using the following equations:

$${\mathrm{MVD}}_{\mathrm{first}}=\left[\left({\mathrm{repetition}}_{\mathrm{last}}-{\mathrm{repetition}}_{\mathrm{first}}\right){/\mathrm{repetition}}_{\mathrm{first}}\right]\times 100$$

$${\mathrm{MVD}}_{\mathrm{fastest}}=\left[\left({\mathrm{repetition}}_{\mathrm{last}}-{\mathrm{repetition}}_{\mathrm{fastest}}\right){/\mathrm{repetition}}_{\mathrm{fastest}}\right]\times 100.$$

$${\mathrm{MVM}}_{\mathrm{first}}=100-\left[\left(\mathrm{mean}-{\mathrm{repetition}}_{\mathrm{first}}\right){/\mathrm{repetition}}_{\mathrm{first}}\right]\times \left(-100\right)$$

$${\mathrm{MVM}}_{\mathrm{fastest}}=100-\left[\left(\mathrm{mean}-{\mathrm{repetition}}_{\mathrm{fastest}}\right){/\mathrm{repetition}}_{\mathrm{fastest}}\right]\times \left(-100\right)$$

Statistical Analysis

All data were distributed normally as determined by the Shapiro-Wilk test of normality (P > 0.05). Descriptive data are presented as means and standard deviations unless otherwise stated. An individual 3-way (deadlift condition [3] × set number [4] × reference repetition [2]) repeated measures analysis of variance (ANOVA) was used to examine the differences between DLnn, DLwn, and DLww in MVD and MVM. A 2-way (deadlift condition [3] × repetition [4]) repeated measures ANOVA was used to investigate differences between DLnn, DLwn, and DLww within the collapsed repetitions for MV. When significant main effects or interactions were obtained, a Holm’s Sequential Bonferroni follow-up test was performed to control for type I error and assess pairwise comparisons. Furthermore, Hedge’s g effect sizes with 95% CI were used and were interpreted as small (0.20-0.49), moderate (0.50-0.80), or large (>0.80). ² To avoid an unnecessarily large number of effect sizes, only moderate and large effects are reported and discussed. All analyses were performed using the R language and environment for statistical computing (version 3.6.3, The R Foundation for Statistical Computing).

Results

Significant interactions and main effects are described in the text, whereas individual P values and effect sizes are illustrated in Figures 1 to 5.

Figure 1.

Condition vs MVD.

Means and standard deviations for MVD using the first (gray circles and lines) and fastest (black circles and lines) repetition in the set as the reference repetitions across all sets. 1RM. 1-repetition maximum; DLnn, deadlifts performed without lifting straps with the load based on without straps 1RM; DLwn, deadlifts performed with lifting straps with the load based on without straps 1RM; DLww, deadlifts performed with lifting straps with the load based on with straps 1RM; MVD, mean velocity decline.

*Significantly greater decline than when using the first repetition in the set as the reference repetition (P < 0.05).

**Significantly greater decline than when using the first repetition in the set as the reference repetition (P < 0.01).

^#Moderate effect size (magnitude of difference, g = 0.50-0.80).

^##Large effect size (magnitude of difference, g > 0.80).

Figure 5.

Means and standard errors for MV during DLwn (black triangles and a dashed line), DLwn (black circles and a dotted line), and DLww (black squares and a solid line) across all sets.

1RM, 1-repetition maximum; DLnn, deadlifts performed without lifting straps with the load based on without straps 1RM; DLwn, deadlifts performed with lifting straps with the load based on without straps 1RM; DLww, deadlifts performed with lifting straps with the load based on with straps 1RM; MV, mean velocity.

**Significantly greater than DLww (P < 0.01).

*** Significantly greater than DLnn (P < 0.05) and DLww (P < 0.01).

^##Large effect size (magnitude of difference, g > 0.8) when compared with both DLnn and DLwn.

Mean Velocity Decline

None of the interactions reached statistical significance: deadlift condition × set number (F = 2.23; P = 0.07), deadlift condition × reference repetition (F = 1.53; P = 0.23), set number × reference repetition (F = 0.90; P = 0.41), and deadlift condition × set number × reference repetition (F = 0.74; P = 0.56). However, significant main effects of deadlift condition (F = 7.52; P = 0.00) and reference repetition (F = 41.61; P < 0.01) were observed due to a significantly lower MVD (I) for DLwn and DLnn compared with DLww, and (II) when using the first repetition of the set as the reference repetition compared with using the fastest repetition of the set (Figures 1 and 2).

Figure 2.

Standardized mean differences with 95% CIs for MVD using the fastest repetition in the set as the reference repetition between the deadlifts performed with and without lifting straps.

Dark gray shade represents small effect size (g = 0.20-0.49), medium dark gray shade represents moderate effect size (g = 0.50-0.79), and light gray shade represents large effect size (g > 0.80). 1RM, 1-repetition maximum; DLnn, deadlifts performed without lifting straps with the load based on without straps 1RM; DLwn, deadlifts performed with lifting straps with the load based on without straps 1RM; DLww, deadlifts performed with lifting straps with the load based on with straps 1RM.

Mean Velocity Maintenance

Similar to MVD, no significant interactions were observed for MVM: deadlift condition × set number (F = 1.31; P = 0.27), deadlift condition × reference repetition (F = 1.50; P = 0.24), set number × reference repetition (F = 0.81; P = 0.45), and deadlift condition × set number × reference repetition (F = 0.74; P = 0.57). However, significant main effects of deadlift condition (F = 5.09; P = 0.02) and reference repetition (F = 42.13; P < 0.01) were observed due to a significantly greater MVM (1) for DLwn and DLnn compared with DLww, and (2) when using the first repetition of the set as the reference repetition compared with using the fastest repetitions of the set (Figures 3 and 4).

Figure 3.

Means and standard deviations for MVM using the first (gray circles and lines) and fastest (black circles and lines) repetition in the set as the reference repetitions across all sets.

1RM, 1-repetition maximum; DLnn, deadlifts performed without lifting straps with the load based on without straps 1RM; DLwn, deadlifts performed with lifting straps with the load based on without straps 1RM; DLww, deadlifts performed with lifting straps with the load based on with straps 1RM; MVM, mean velocity maintenance.

*Significantly lower maintenance than when using the first repetition in the set as the reference repetition (P < 0.05).

**Significantly lower maintenance than when using the first repetition in the set as the reference repetition (P < 0.01).

^#Moderate effect size (magnitude of difference, g = 0.50-0.79).

^##Large effect size (magnitude of difference, g > 0.80).

Figure 4.

Standardized mean differences with 95% CIs for MVM using the fastest repetition in the set as the reference repetition between the deadlifts performed with and without lifting straps.

Dark gray shade represents small effect size (g = 0.2-0.5), medium dark gray shade represents moderate effect size (g = 0.5-0.79), and light gray shade represents large effect size (g > 0.8). 1RM, 1-repetition maximum; DLnn, deadlifts performed without lifting straps with the load based on without straps 1RM; DLwn, deadlifts performed with lifting straps with the load based on without straps 1RM; DLww, deadlifts performed with lifting straps with the load based on with straps 1RM; MVM, mean velocity maintenance.

Repetition-by-Repetition Analysis

A significant deadlift condition × repetition (F = 3.40; P = 0.02) interaction was observed with DLwn and DLn allowing for greater MV than DLww during all repetitions (P < 0.01) and DLwn allowing for greater MV than DLnn (P = 0.01-0.03) only during repetitions 2 to 4 (Figure 5). Overall, the second repetition of the set was generally the fastest (54.1% of cases), followed by the first repetition (34.4%), third repetition (8.9%), and fourth repetition (2.6%).

Discussion

This study was designed to explore the behavior of MVD and MVM during deadlifts performed with and without lifting straps. The main findings revealed that (1) MVD was significantly greater for DLww compared with DLwn and DLnn, with this result not being affected significantly by the set number or reference repetition; (2) MVM was significantly greater for DLwn and DLnn compared with DLww regardless of the set number or reference repetition; (3) using the first repetition as the reference repetition resulted in less MVD and greater MVM across the sets and deadlift conditions compared with using the fastest repetition; and (4) the second repetition of the set (54.1%), and not the first repetition (34.4%), was generally the fastest repetition across deadlift conditions and sets. Taken together, these results suggest that lifting straps do not affect velocity decline or velocity maintenance when the same absolute load is lifted. Regardless of whether lifting straps are used, using the first repetition as the reference repetition may underestimate velocity decline and overestimate velocity maintenance because it is not consistently the fastest repetition in the set.

Our first hypothesis was supported only partially since significantly less MVD and greater MVM were observed during DLwn compared with DLww but not compared with DLnn. Interestingly, however, a previous study showed that lifting straps can allow for faster absolute velocity compared with no lifting straps during the same RT protocol. ¹⁰ This discrepancy could likely be explained by the variables being analyzed and their sensitivity to a given RT protocol. Namely, the participants in the present study performed only 4 repetitions with 80% of 1RM, whereas previous studies have shown that fatigue increases when more repetitions are performed in the set, which would also result in a greater velocity decline and less velocity maintenance.^25,30 Considering that, and since MVD was relatively low (<17%) across sets and conditions, it seems that the number of repetitions performed in the present study was the main reason for the lack of effect of lifting straps on MVD and MVM. This, coupled with previously reported ergogenic effects of liftings straps on absolute velocity outputs, ¹⁰ suggests that MVD and MVM are more sensitive to the number of repetitions being performed in the set than absolute velocity outputs, which has important training monitoring implications. It is important to note that the highest MVD recorded in this study was ~16%, whereas the lowest MVM recorded was ~93%. Therefore, it is evident that the absolute, maintenance, and decline of lifting velocity may each tell a different story and, hence, should be classified independently. ¹¹ By using all 3 during RT, practitioners will gain a greater insight into (1) overall mechanical performance (absolute velocity), (2) overall capacity to maintain the maximum velocity performance during the whole set (velocity maintenance), and (3) degree of neuromuscular fatigue experienced at the end of the set (velocity decline).

Whether using MVD, MVM, or any other velocity-fatigue measurement, it is important to choose the appropriate reference repetition from which all others are compared with or calculated from. Previous studies have used both the first and the fastest repetition of the set as the reference repetition,^4,15,19,30 but the findings of the present study suggest that this is not a trivial decision since using the first repetition as the reference repetition consistently yielded lower MVD and greater MVM compared with using the fastest repetition. These findings agree with a recent study that showed that, when using the first compared with the fastest repetition as the reference, more repetitions could be performed before exceeding a given velocity loss threshold. ⁷ In addition, previous studies have also shown that the first repetition was not always the fastest,^6,7,16 which was also confirmed in the present study. For instance, while examining deadlift condition × repetition interaction (Figure 5), it can be seen that the second repetition was also the fastest, on average, across the conditions. This is not surprising considering the absence of the stretch-shortening cycle, and its associated potentiation benefits in the first repetition as opposed to the second repetition of the deadlift exercise.^18,20 Therefore, the use of the first repetition as the reference point for MVD and MVM calculations could be problematic when the first repetition is not the fastest since (1) excessive fatigue could be induced when the number of repetitions is prescribed based on velocity loss thresholds and (2) the overall RT-induced fatigue could be underestimated, both of which could compromise recovery and result in unintended neuromuscular stimuli. In that regard, practitioners should consider the mechanics of the training exercise (eg, involvement of the stretch-shortening cycle, load lifted, or inter-repetition rest),^6,16,20,29 which is likely to influence the coexistence of fatigue and potentiation, and the reference repetition from which all others are compared with or calculated from to ensure precise monitoring of the RT-induced fatigue using MVD and MVM.

Several methodological aspects should be considered when interpreting our findings. First, the magnitude of velocity decline may differ between exercises. ¹ Therefore, it is possible that the effects of lifting straps on velocity decline and maintenance could also differ in other pulling exercises where grip strength may play a larger role. Second, velocity decline and maintenance variables could be sensitive to the number of repetitions performed in the set, making it plausible that lifting straps could indeed have a positive effect on velocity decline and maintenance when more repetitions are performed in the set, or when the sets are performed closer to failure. Third, our findings are representative only of a resistance-trained male population who could deadlift between 150% and 230% of their body mass for at least 1 repetition without lifting straps. Finally, our findings do not necessarily transfer to other pulling exercises where lifting straps are commonly used.

Conclusion

Analyzing the absolute mean and peak values of a given mechanical variable is not the same as assessing the change in performance (ie, decline and maintenance) throughout an RT session. This has been evidenced in the present study since positive effects of using lifting straps on velocity decline and maintenance were not observed even though previous research showed positive effects of lifting straps on absolute velocity outputs. Regardless of whether lifting straps are used, the use of the first repetition as the reference repetition underestimated velocity decline and overestimated velocity maintenance because the second repetition was the fastest in more than half of the sets.

Practical applications

The present study shows that the use of lifting straps does not affect velocity decline and maintenance during a deadlift exercise despite previous work showing faster absolute velocities with lifting straps compared with the same load without lifting straps. Therefore, although lifting straps may allow for greater absolute velocities at a given load, the straps do not affect the degree of relative velocity change. Furthermore, since many factors such as stretch-shortening cycle, force potentiation from a previous muscle contraction, load lifted, and the use of inter-repetition rest periods can all affect performance of the first repetition in the set, the use of the first repetition as the reference repetition might lead to underestimated velocity decline and overestimated velocity maintenance when using these variables to quantify RT-induced fatigue or effectiveness of different training protocols. Therefore, practitioners should consider these factors and adopt a more holistic evaluation of the RT protocols using absolute, decline, and maintenance of velocity outputs to ensure precise monitoring of the RT-induced fatigue.