Surprisingly, the controversy of whether deadlifts should be substituted with squat exercises during training continues to resonate throughout the powerlifting community. Many individuals involved in the sport of powerlifting believe that the squat and deadlift have such similar lifting characteristics that the lifts will yield comparable training results, with a cross-over effect. Numerous coaches and competitors mistakenly continue to substitute various squat exercises into their training programs in place of the deadlift in hopes of miraculous performance improvements in the deadlift. However, the resemblance appears to be primarily restricted to submaximal deadlift loads typically used during the early stages of a training program (10). Even though the squat and deadlift appear similar from a simplistic mechanical standpoint, the lifts are considered quite different when analyzed under maximal loading conditions, as used in powerlifting competitions. It is extremely difficult for an individual to identify the minute movement pattern differences visually, but it is very easy to isolate variations using sophisticated motion analysis instrumentation. From a biomechanical perspective, the squat and deadlift are much more different from one another than currently believed by many in the powerlifting community.
Two basic strategies exist for lifting weights off of the ground (i.e., the deadlift) (9). One strategy is referred to as the leg-lift, which displays a flexed knee position with a relatively vertical trunk position. The method demonstrates a movement that exhibits simultaneous and relatively equal angular changes about the hip, knee, and ankle that closely resembles the back squat movement. Conversely, the back-lift strategy exhibits an extended knee and flexed trunk position during the initial phase of the movement, indicating a significantly increased rate of knee extension. The knee movement coupled with heavy weight lifting causes the hips to rise rapidly, thus creating an increase in trunk lean. The leg-lift method demonstrates a reduced load on the lumbar spine, which is considered safer, but the knees are heavily loaded, whereas the back-lift style subjects the lumbar region to extremely high resultant forces and moments. Depending on the lifting strategy, lower-extremity muscle groups are recruited in either a synergistic (leg-lift) or sequential (back-lift) manner to generate the necessary muscle moments. Bejjani et al. (2) reported that the leg-lift method approaches the straight leg technique when lifting extremely heavy weights caused, in part, by biomechanical constraints and the inability to maintain lumbar lordosis. During heavy weight lifting, the leg-lift method does not appear to be the preferred strategy and may not be even possible (18-20). In summary, no one tends to use one method explicitly, but rather people use a combination of the 2 different styles. However, when an individuals near their maximal deadlift capacity, the back-lift strategy appears to be the most commonly used technique.
To perform a comparative analysis, a common phenomenon specific to each of the individual lift types was needed to serve as a marker for distinguishing differences between the lifts. The event, described as the “sticking point,” was selected because it has been reported to occur in various weight lifting movements, including the back squat and the deadlift (6,15,16). The sticking point of a movement is where the upward momentum of the barbell is momentarily decreased or stopped. This event is particularly important because the success in weight lifting depends on a continuous motion of the barbell past this region of the lift. By isolating the sticking point and identifying the biomechanical constraints associated with this region, one could provide training implications that address lifting technique and the possible implementation of specific assistance-type exercises.
Numerous weight lifting analyses have focused on different squat techniques (1,5,7,12,14) or different deadlift styles (3,4,8) to determine which method is best for strength development or lower-extremity rehabilitation (8,21), but a comparative analysis between lift types has never been reported. Investigating the kinematics of both lift types concurrently while performed by a common subject pool could produce information that may challenge current training philosophies.
Experimental Approach to the Problem
The objective of the study was to compare the back squat and the conventional style deadlift to determine whether a biomechanical relationship exists between the individual lift types. A competition setting was selected over a laboratory-based analysis because of the different lifting strategies used, submaximal versus maximal loads. An all-out effort in competition as lifters attempt to set personal or state/national records is very difficult to replicate in a laboratory. For this reason, a competition environment was deemed best for addressing the issue of training specificity in relation to the squat and deadlift cross-over effect.
The conventional style deadlift was selected for 2 reasons: a) fewer lifters use the sumo style deadlift in competition (small subject pool), and b) the majority of the sumo style deadlifters do not perform the back squat with a wide stance, so the significantly different foot placement would affect the biomechanical comparison. Only lifters that used a conventional-style deadlift and relatively narrow squat stance, heels approximately shoulder width, were selected for analysis. Similar foot placement was an important control factor for comparing the individual lift types.
The 25 male participants were competitors in a regional powerlifting contest, which also served as a national qualifier. The competition attracted lifters with varying skill levels, which provided a heterogeneous group of participants. Only lifters who successfully completed a squat and deadlift were randomly selected for analysis at the conclusion of the contest. Before competition, body weight (93.63 ± 30.9 kg), body height (1.82 ± 0.15 m), and age (28 ± 8 yr) were recorded after receipt of a signed consent from each participant. Before data collection, the research project was reviewed and approved by the Institutional Review Board at Georgia State University.
Four synchronized video cameras (60 Hz) were used to quantify the 3-dimensional analysis. The cameras were oriented to avoid any possibility of the spotters, judges, and weight plates obstructing the camera's view of the competitors (i.e., body landmarks). A peak performance video/analog motion measurement system was used to extract kinematic data from the conditioned video signals (17), and then data were smoothed using a Butterworth digital filter (cutoff frequency 6 Hz). Four bilateral joint centers (shoulder, hip, knee, and ankle) and 2 podiatric landmarks (lateral aspect of the calcaneus and lateral aspect of the 5th metatarsal) were manually digitized. In addition, hand and bar contact points were included for the deadlift to accurately measure bar velocity. Also, the barbell was assumed to be fixed at the shoulder during the squat, so the shoulder joint markers were used to quantify bar motion. Contrary with previous studies, the barbell end was not selected as a point for analysis because bar flex (oscillation) occurs throughout the execution of the lifts. Three stages were identified during the deadlift: a) lift-off (LO); 2) knee passing (KP); and 3) lift completion (LC). In accordance with the United States Powerlifting Federation (USPF) guidelines, the competitors were allowed 3 lift attempts for each of the individual lifts (11). The maximal load successfully lifted, for each lift type, was selected for analysis of each of the lifters.
Means ± SD were calculated for each of the biomechanical parameters. Paired (samples) t-tests were used to detect significant differences in the kinematic mean scores for the different lift types (α = 0.01). Calculated p values lower than 0.01 were considered evidence for statistical significance.
Absolute and relative angles were calculated geometrically along with the 3-dimensional linear and angular kinematics of the following parameters: bar, shoulder, hip; knee, and ankle. It was observed that, during the deadlift, certain body segments ascended before barbell LO, so the “ascent phase” was defined as the upward movement of the barbell, unless otherwise stated.
Vertical Bar Velocity (Sticking Point)
Each kinematic variable was reported at 3 different points: P1 signified the start of the ascent phase; P2 represents the first point where vertical bar momentum begins to diminish; and P3 represents the first point of pronounced acceleration (sticking point) after the previous phase. The sticking point is an important position because if the lifter is unable to overcome the inertia of the lifter-barbell system and increase vertical velocity at P3, the lift will not be successful.
Vertical bar velocity during the ascent phase was used to identify the sticking region. Figures 1 and 2 illustrate the sticking point associated with the squat and the conventional style deadlift. A significant difference (p < 0.01) exists between the squat and deadlift at P1 but not at P2 and P3.
Each parameter selected for analysis indicated significantly different (p < 0.01) relative angular positions at the selected points. The hip and knee angular displacements for the squat displayed relatively equal changes throughout the sticking point region at 14.45° and 10.25°, respectively.
The trunk demonstrated a decrease in angular position (2.66°) during the sticking region of the deadlift, whereas hip extension increased (10.25°). Similarly, during the squat, the trunk displayed minimal angular change (1.11°), whereas the hip exhibited 14.45° of extension. The absolute thigh angle at the sticking point for the squat and deadlift were 32.54° and 57.42°, respectively.
Knee Angle Versus Hip angle
The deadlift showed 3 distinct or segmented phases defined by a dominant joint action: LO, knee extension; KP, hip extension; and LC, knee/hip extension. The squat lift produced a linear relationship between the joint movements, illustrating a more synergistic or simultaneous movement.
These data were used in a quantitative manner to substantiate how segmental lengths may influence and alter moment arm lengths between the different lift types. Only trunk length was significantly different (p < 0.01) between the lift types.
A more detailed kinematic model would have been preferred, but the USPF regulations would not permit the use of reflective markers during competition, which could have potentially impeded the performance of the lifters. However, the powerlifting championship setting provided an ideal opportunity to analyze individuals performing the squat and deadlift while lifting extremely heavy loads. Even though the lifts demonstrate fundamental differences and use different lifting strategies, a detailed analysis was used to address a common misconception that a cross-over effect exists between the lifts. The most obvious differences between the squat and deadlift are the placement of the barbell and the lifter's body position during the execution of the lifts; however, the kinematic differences between the lifts are not as evident without motion analysis instrumentation. The deadlift, unlike the back squat, consists only of an “upward” movement when performed in competition, so an analysis of the ascent phase was selected for 2 primary reasons: a) it was common between lift types and b) it contained the sticking point for both movements.
The statistical analysis revealed significant differences between the 2 lifts for the selected kinematic parameters (p < 0.01). To identify the sticking point for the squat and deadlift, bar velocity was selected because it best represents the performance and outcome of the overall lift (Table 1). It was shown that the sticking point of the squat (Figure 1) and deadlift (Figure 2) not only occurred at different hip, knee, and ankle angular positions but also displayed different bar velocities. These findings support the argument that the sticking point mechanisms differ between the squat and the conventional- style deadlift caused, in part, by biomechanical and physiologic factors affecting muscle force production. Some of the variability among the kinematic parameters could be attributed to the differing skill levels or differences in antropometric measurements between the individual lifters. Further research in this area should investigate how high-skill level competitors negotiate the sticking point region as compared with lesser-skilled lifters.
The back squat exercise is considered paradoxical because of the cocontracting antagonistic muscles of the lower extremities during the execution phase. Lombard and Abbott (13) described the phenomenon involving bi-articulate muscles in a manner that joint movement is determined by the dominant muscle moment about that particular joint regardless of the muscle's multifunctional role. The relative knee and hip angular displacements during the back squat movement display similar rates of change throughout the “sticking point” region (Table 2). Figure 3 depicts a relatively linear relationship between the knee and hip throughout the ascent phase of the squat, a condition during which cocontraction of the rectus femoris and the hamstring group (biceps femoris, semitendenosus, and semimembranosus) is evident. In this case, the extensor muscle moment generated at the hip by the hamstrings is in excess of the flexor muscle moment of the rectus femoris. Similarly, at the knee, the extensor moment of the quadriceps dominates the flexor muscle moment of the hamstring group (19,22). Bar velocities quantified during the ascent phase of the squat implies the net extensor moment is diminished at the “sticking point” region (Figure 1). In addition, the mean absolute knee angle (Table 3) reported at the sticking point is similar to past research findings (16).
Conversely, the deadlift exhibits different lifting mechanics as compared with the squat, with the most obvious being the 3 distinct stages (Figure 4) identified during the execution or ascent phase: LO, KP, and LC. Interestingly, these independent phases have never been clearly defined in previous research, and the availability of this information could challenge the current training methods that target the sticking point in the deadlift. The different phases were defined based on the kinematics and further identified by the barbell location relative to anatomic landmarks. The LO phase begins with the weight on the floor and proceeds until the barbell nears the tibial tuberosity (approximately 6 cm distal to the patella). The movement exhibited predominant knee extension throughout the entire phase (Table 2), and the trunk angle actually decreased, indicating an increase in trunk lean during the initial phase of the lift (Table 3). The second stage was defined as the distance between the tibial tuberosity to a point approximately 6 cm proximal to the patella. During this phase of the lift, the movement was dominated by hip extension with relatively minimal knee extension. An important note, the KP stage includes the sticking point position, which is located near the inferior portion of the patella. The final stage of the deadlift (LC) transitions from the KP phase until the body is standing completely erect. The LC portion of the lift demonstrated a combination of both hip and knee extension. The deadlift does not necessarily follow the Lombard model because the execution of the lift is partitioned into 3 distinct phases.
The segmental trunk length was defined as the distance between the hip and shoulder joint centers. The relative trunk lengths differed (p < 0.01) between the squat and deadlift (Table 4), suggesting the lifters demonstrated different postural positions during the execution phase of the lifts. During the back squat, lifters were able to maintain lumbar lordosis with a slightly arched but rigid spinal column. However, the deadlift exhibited a different trunk configuration, which suggests that under maximal load conditions, the spinal column displays an abnormal curvature of the spine during the execution of the deadlift. Specifically, the trunk was unable to maintain lumbar lordosis, and a prominent kyphotic condition was evident at the thoracic region of the spinal column, which produced a rounded back posture. It has been reported that maximal and submaximal deadlifts demonstrate different lifting techniques and require different lifting strategies (10).
In summary, the sticking point for both the squat and deadlift identified the portion of the lift where the involved muscle groups are disadvantaged, possibly because of anatomic (bone architecture and alignment) and mechanical (resultant muscle forces and moments) factors specific to that region. The “sticking point” is a position common to each lift but occurred at 2 distinctly different locations during the ascent. Three absolute conclusions are drawn from the data, supporting reasons why the lifts are different: a) evidence indicates that the squat represents a synergistic or simultaneous movement, whereas the deadlift demonstrates a sequential or segmented movement; b) both lifts demonstrate different sticking point locations, indicating different lifting strategies supported by angular position and kinematic measurements; and c) the squat and deadlift exhibited different trunk configurations, which had a direct influence on the lifting techniques under maximal load conditions. The kinematic analysis of the squat and the conventional deadlift indicate that the individual lifts are markedly different, implying that no direct or specific cross-over effect exists between the individual lifts. Furthermore, insight into possible causes for the noted drop in vertical bar velocity at the key positions during the ascent cannot be clearly gained from the results of this study. Because of the complicated musculoskeletal mechanics of the numerous bi-articulate muscles involved in the execution of the squat and conventional style deadlift, interpretations of the resultant forces and moments are difficult. Although the kinematic data reaffirmed that there may be changes in muscle action, mechanical disadvantages, muscle angle of pull, and muscle length-tension relationships near this part of the movement, a biomechanical dynamic model is needed before the musculoskeletal dynamics can be explained adequately. The scope of this analysis was to determine gross kinematic differences between the back squat and the conventional style deadlift.
Currently, a large portion of the powerlifting community assumes that the sticking point for the deadlift occurs approximately at knee level. This assumption was supported by the findings of this study; however, momentum began to diminish before the sticking point. A popular exercise for training the sticking point of the deadlift is referred to as a “lockout” or “partial deadlift” performed using a “power rack.” Typically, lifters place the barbell approximately at knee level and proceed to lift the bar from this starting position until the hips and knees are fully extended. In retrospect, as the results suggest, starting the partial deadlift at the knees may be incorrect because the findings from the analysis indicated that vertical bar momentum began to decrease before the bar reached the knee position. Therefore, in contradiction with current training methods, the barbell should be positioned approximately 6 cm below the knees to target the entire region, signifying a decrease in momentum including the sticking point. The problem most people encounter while performing lockouts is the inability to replicate body position specific to the deadlift at that point of the lift. Performing lockouts improperly on a regular basis could potentially have a negative affect on the deadlift technique. Also, poor technique may inhibit the recruitment process of the involved musculature in regard to the biomechanical factors affecting muscle force production. Contrary with popular belief, deadlifts should not be eliminated from training programs and replaced with squat exercises. Rather, the deadlift should be properly and strategically placed in a periodized training regimen to maximize performance results. In conclusion, on the basis of the findings from this analysis, the best way to improve the deadlift is to deadlift.
1. Anderson, R, Courtney, C, and Carmell, E. EMG analysis of the vastus medialis/vastus lateralis muscles utilizing the unloaded narrow and wide-stance squats. J Sport Rehabil
7: 236-247, 1998.
2. Bejjani, FT, Gross, CM, and Pugh, JW. Model for static lifting, relationship of loads on the spine and the knee. J Biomech
17: 281-286, 1984.
3. Brown, EW and Abani, K. Kinematic and kinetics of the dead lift in adolescent power lifters. Med Sci Sports Exerc
17: 554-563,. 1985.
4. Cholewicki, J, McGill, S, and Norman, R. Lumbar spine loads during the lifting of extremely Heavy Weights. Med Sci Sports Exerc
23: 1179-1186, 1991.
5. Ebben, WP, Leigh, DH, and Jensen, RL. The role of the back squat as a hamstring training stimulus. Strength Cond J
22: 15-17, 2000.
6. Elliot, BC, Wilson, GJ, and Kerr, GK. A biomechanical analysis of the sticking point
in the bench press. Med Sci Sports Exerc
21: 450-462, 1989.
7. Escamilla, RF. Knee biomechanics
of the dynamic squat exercise. Med Sci Sports Exerc
33: 127-141, 2001.
8. Escamilla, RF, Fransisco, AC, Kayes, AV, Speer, KP, and Moorman, CT. An electromyographic analysis of sumo and conventional style deadlifts. Med Sci Sports Exerc
34: 682-688, 2002.
9. Freivalds, A, Chaffin, DB, Garg, A, and Lee, K. A dynamic biomechanical evaluation of lifting maximum acceptable loads. J Biomech
17: 251-262, 1984.
10. Gracovetsky, S and Farfan, H. The optimum spine. Spine
11: 543-571, 1984.
11. International Powerlifting Federation Guidelines, 2008.
12. Lander, JE, Bates, BT, and Devita, P. Biomechanics
of the squat exercise using a modified center of mass bar. Med Sci Sports Exerc
18: 469-478, 1986.
13. Lombard, WP and Abbott, FM. The mechanical effects produced by the contraction of individual muscles of the thigh of the frog. Am J Physiol
20: 1-60, 1907.
14. McCaw, ST and Melrose, DR. Stance width and bar load effects on leg muscle activity during the parallel squat. Med Sci Sports Exerc
31: 428-436, 1999.
15. McGuigan, MR and Wilson, BD. Biomechanical analysis of the dead lift. J Natl Strength Cond Res
10: 250-255, 1996.
16. McLaughlin, TM, Dillman, CJ, and Lardner, TJ. A kinematic model of performance in the parallel squat by champion power lifters. Med Sci Sports Exerc
9: 128-133, 1977.
17. Peak Performance Technologies, Inc.Inc. User's Reference Guide
18. Rasch, PJ and Burke, RK. Kinesiology and Applied Anatomy
(6th ed). Philadelphia: Lea and Febiger, 1978.
19. Schipplein, OD, Trafimow, JH, Andersson, GBJ, and Andriacchi, TP. Relationship between moments at the L5/S1 level, hip and knee joint when lifting. J Biomech
27: 907-912, 1990.
20. Schultz, AB and Andersson, GBJ. Analysis of loads on the lumbar spine. Spine
6: 76-82, 1981.
21. Wright, GA. Electromyographic activity of the hamstrings during performance of the leg curl, stiff-leg deadlift, and back squat movements. J Strength Cond Res
13: 168-174, 1999.
22. Zajac, FE and Gordon, ME. Determining muscle's force and action in multi-articular movement. Exerc Sport Sci Rev
17: 187-230, 1989.
Keywords:© 2009 National Strength and Conditioning Association
biomechanics; sticking point; sequential; synergistic; partial deadlift