owerlifting is a sport made up of three events, the squat,
bench press and deadlift. For each event, the ultimate goal
is to lift as much weight as is possible. In the deadlift, a lifter
lifts the barbell off of the floor until standing upright. The lift
is finished upon extending the knees and hips with scapula
retracted. Two styles of the deadlift are used in competition.
The sumo style uses a wide foot stance, upright posture, and a
grip width that is narrower than the feet.
conventional style deadlift uses a narrow foot stance, generally
a more bent-over posture, and a grip outside of the legs.
Three key phases have been identified in the literature for
the conventional deadlift.
The first phase, or lift-off, occurs
when the lifter first applies force to the bar and the bar rises off
of the floor. The second phase, knee passing, occurs when the
bar moves from below to above the knee. The third phase, or
lift completion, occurs when the lifter transitions into a full
upright position. While these specific regions of the deadlift
are known, little has been done to examine how each position
might contribute to deadlift performance. The most
disadvantageous position represents a limiting factor in overall
performance, thus identification of this position may lead to
better training prescriptions.
To the authors’ knowledge, no literature exists that assesses
the force generation capabilities of lifters in these phases of the
deadlift, however in a number of studies examining the
isometric mid thigh pull (MTP), a weightlifting-specific
position, a variety of athletes produced high levels of peak
Peak force measured in these studies showed
moderate to strong relationships with dynamic mid-thigh pulls,
jumps and a number of other dynamic measures. Therefore,
since little is known about the force generation capabilities of
lifters in the key phases of the deadlift, the purpose of the
study was to evaluate the isometric maximum strength of
powerlifters in bar positions corresponding to key phases of
the deadlift and also to compare those positions to the MTP,
given the strong relationship the MTP shares with a variety of
Experimental Approach to the Problem
Data obtained in an athlete monitoring program were
assessed using a repeated measures design to assess peak force
production differences between key positions of the deadlift. A
repeated measures ANOVA and paired t-tests were used to
assess force differences between positions.
Fourteen competitive powerlifters who could deadlift a
minimum of 2.5 x body mass (BdM) using the conventional
style using only a belt or competed regularly volunteered for
this investigation. Based upon training history questionnaires
all subjects reported that they did not regularly perform
weightlifting movements or their variants. Some lifters
reported using the sumo style most often in competition (n=4),
Isometric Strength of Powerlifters in Key Positions
of the Conventional Deadlift
George K. Beckham, Hugh S. Lamont, Kimitake Sato, Michael W. Ramsey, G. Gregory Haff, Michael H. Stone
Objectives: To determine if force differences exist between isometric pulling positions corresponding to key positions of the
Design: Cross-sectional evaluation of isometric strength
Methods: 14 powerlifters performed isometric pulls on a force plate at 3 key positions related to the deadlift (at the floor,
just above the patella, and 5-6 cm short of lockout) and in the mid thigh pull position (MTP). A 1x4 repeated measures
ANOVA was used to ascertain differences between the various pulling positions tested. Bonferroni-adjusted paired
samples t-tests were used post-hoc.
Results: Forces generated at each bar height were significantly different (F(3,39) = 51.058, p<0.05, η
samples t-tests showed significant differences between positions, revealing a trend of greater force generation at
increasing heights for positions corresponding to the deadlift. Force generated in the mid thigh pull position was
significantly higher than any other position.
Conclusion: In positions corresponding to the deadlift, force generation increases at higher bar heights.
(Journal of Trainology 2012;1:32-35)
Key words: powerlifting
isometric mid thigh pull
Received October 17, 2012; accepted November 12, 2012
From Center of Excellence for Sport Science and Coach Education, Department of Kinesiology, Leisure, and Sports Science, East Tennessee State
University, Johnson City, TN, USA (G.K.B., H.S.L., K.S., M.W.R., M.H.S), and Centre for Exercise and Sport Science Research, Edith Cowan University,
Perth, Australia (G.G.H., M.H.S).
Communicated by Takashi Abe, PhD
Correspondence to Mr. George K. Beckham, East Tennessee State University, PO Box 70654. Email: firstname.lastname@example.org
Journal of Trainology 2012:1:32-35 ©2012 The Active Aging Research Center http://trainology.org/
Beckham et al. Isometric Strength of Powerlifters in Key Positions of the Conventional Deadlift 33
but all lifters reported training regularly using the conventional
style (n=14, age range: 18-39, height: 178.6±9.8cm, BdM:
109.9±20kg, conventional deadlift 1-RM: 248.5±18kg). Each
subject was screened by questionnaire for injury prior to
testing. Athletes were informed of all testing procedures and
possible risks, and voluntarily signed an informed consent
document as outlined by University Institutional Review
The warm-up routine was a standardized protocol with a
small amount of possible modification (within the specified
range) to more closely match the typical warm-up routine of
the lifter. Warm-ups were as follows: 2-5 repetitions at 35% of
1-RM, followed by 90 seconds rest, 2-3 repetitions at 50%
1-RM, followed by 120 seconds rest, 1-2 repetitions at 65%
1-RM, followed by 150 seconds rest, then 1 repetition at 75%
1-RM, followed by 180 seconds rest. Warm-up loads were
determined using the athletes’ belt-only conventional personal
Isometric Testing Procedures
All isometric testing was completed in a custom designed
power rack that allows fixation at any height. Athletes stood
on a 91.4 x 91.4 cm force plate (Rice Lake Weighing Systems,
Rice Lake, WI) to measure vertical ground reaction forces. Bar
heights for each testing condition were chosen to correspond
to the three key positions achieved in the deadlift and the
isometric mid-thigh pull. For the first height, the center of the
bar was placed at 22.5 cm from the floor to correspond to the
position of the barbell in the start of the deadlift. The second
bar position was placed immediately superior to the patella
from standing. The third corresponded to the same body
position as used in the MTP.
The fourth position used the
same bar height as the third, but with a self-selected body
position corresponding to one that would be achieved in a
deadlift. Pilot testing indicated that the fourth height results in
a body position with the bar 4-6 cm from deadlift lockout.
Intra-session test-retest reliability (intraclass correlation,
coefficient of variation, respectively with 90% CI’s for each)
of peak force for each position was excellent: floor: 0.99
(0.98-1.0), 1.2% (0.9%-1.8%), knee: 0.98 (0.96-0.99), 2.0%
(1.5%-2.9%), IMTP: 0.92 (0.80-0.96), 5.0% (3.8%-7.5%),
lockout: 0.88 (0.70-0.94), 4.6% (3.5%-6.9%).
Each condition was performed in order, 1-4, to maintain
standardization among athletes and result in a uniform fatigue.
Pilot testing indicated that forces and perceived difficulty
increased as the athletes used the higher bar positions, thus the
order was chosen to correspond to what was likely least
fatiguing to most fatiguing. The conditions were separated by
10 minutes of rest, during which time the athletes remained
seated. Athletes were secured to the bar using lifting straps and
athletic tape. Each subject assumed the position he would be
using for the pull condition, and once body position was
stabilized (verified by visual monitoring of both the athlete
and force trace), the athlete was given a countdown. Nominal
pre-tension was allowed to minimize slack in the subject’s
body prior to the pull (monitored by force-trace and instruction
to the lifter) to ensure that little or no vertical acceleration of
the athlete occurred. The subject performed two warm-up
attempts separated by 90-120 seconds, each at a subject-
estimated 50% and 75% of maximum. The athletes then
performed 2 to 3 maximal attempts for 3-4 seconds each,
separated by 2-3 minutes. The attempt was terminated when a
plateau or consistent decrease in force was observed. A third
trial was only performed if a ≥250N difference in PF was
observed between trials, a countermovement was observed, or
if the athlete did not follow directions.
The highest observed force from each pull obtained using a
custom analysis program (National Instruments, Austin, TX)
was designated peak force (PF). PF measurements from both
trials were averaged. Peak force was allometrically scaled
(APF) using the equation [y=result∙BdM
Analog data from the force plate were amplified and
conditioned (low-pass at 16 Hz; Transducer Techniques,
Temecula, California). An AD converter (DAQCard-6063E,
National Instruments, Austin, TX) allowed for collection at
1000 Hz and further low-pass filtering using a software-based
Order Butterworth filter at 100 Hz.
For the purpose of comparing kinetic measures at each of
the four pulling positions, a repeated measures ANOVA (RM
ANOVA) was used for each dependent variable considered,
using Bonferroni adjusted paired t-tests (p=0.008) for post-
hoc analysis. RM ANOVA and post-hoc tests were performed
for unscaled and allometrically scaled force. Alpha was
designated at p=0.05. All statistical analysis was performed
using SAS 9.2 (Statistical Analysis System, SAS Institute Inc.,
Cary, NC). Effect sizes were evaluated with the method of
PF and APF measures can be found in Table 1. There was a
significant main effect for PF (F(3,39) = 87.44, p<0.0001),
of 0.871. APF measures were significant for main
effect (F(3,39) = 88.23, p<0.05) with η
of 0.872. Results of
paired t-tests can be found in Table 1. Effect sizes of paired
t-tests for PF and APF were as follows: floor vs. knee, 1.50,
1.97; floor vs. MTP 3.66, 4.22; floor vs lockout 3.04, 3.08;
knee vs. MTP 2.10, 2.80; knee vs. lockout 1.40, 1.52; MTP vs.
lockout 1.23, 1.27.
Athletes produced different PFs at each position (floor, knee,
MTP, lockout). The changing bar height resulted in different
body positions for each pull, and thus a differing ability to
apply force. Interestingly, positions directly related to deadlift
performance (floor, knee, lockout) tended to increase force in
the higher bar positions. PF and APF in the floor position were
significantly less than both the knee and lockout positions
(effect size of large to very large). There was also a significant
Journal of Trainology 2012;1:32-3534
difference between knee and lockout positions (with large
effect size). This finding may be due to a number of reasons.
Brown & Abani
found that horizontal hip moment to the bar
center of mass (COM) decreased with higher bar positions in
the deadlift. While Escamilla et al.
did not test for significant
differences, they reported a trend of decreasing horizontal
moment arm to the barbell COM at the ankle, hip and knee as
the lifter ascended from lift-off to knee passing. This decreased
moment at higher positions may allow for better mechanical
advantage at the hip, thus increasing the resultant generation of
upward force on the bar.
One confounding issue is the fact that two studies have
found that the sticking region occurs at a point roughly around
Because biomechanical disadvantage causes the
sticking point to occur at a certain range of motion, the total
force generating capability at that position should be reduced
(net extensor moment and force applied to the bar). Therefore,
based on the two aforementioned studies, the force generating
capabilities of deadlifters at the knee position should be less
than the floor position, not more, as was found in the present
study. It is possible that anthropometric characteristics
predispose one to certain sticking points, but no known
research exists to assert this. Another possibility is that in the
Escamilla et al.
and Hales et al.
studies the lifters were using
a powerlifting deadlift suit. If this was the case, then the
sticking regions of the lifts may be higher due to the assistance
afforded the lifter by lifting suits.
Also possible is that the
position used by athletes in the present study is different than
what athletes use in a maximal deadlift. If the isometric pull
allows for a more ideal body position than is attained during
the deadlift, greater forces might be achieved, thus
representing a possible limitation of this study. Further
research should confirm this.
It is interesting that the MTP position allowed the lifters to
produce the greatest amount of force, despite the lockout
position being more similar in position to the deadlift. The
lifters generally performed well in the lockout position
(understandable given that they regularly train a movement
that requires it, i.e. the deadlift, and do not regularly train in
the MTP position); therefore the MTP position must provide a
substantial mechanical advantage that overcomes even the
frequent training in the deadlift-specific position. The greater
Table 1. Results of isometric testing and paired t-tests
Measure Position Mean ± SD Significance
Peak Force (N)
Floor 3380.0 ± 377.0 †, ‡, §
Knee 4093.0 ± 559.0 *, ‡, §
Mid-Thigh Pull 5829.0 ± 867.0 *, †, §
Lockout 4910.0 ± 605.0 *, †, ‡
Peak Force (N.kg
Floor 148.5 ± 12.7 †, ‡, §
Knee 179.8 ± 18.6 *, ‡, §
Mid-Thigh Pull 256.4 ± 33.9 *, †, §
Lockout 216.6 ± 28.6 *, †, ‡
* = significantly different than floor position p <0.001
† = significantly different than knee position p <0.001
‡ = significantly different than MTP position p <0.001
§ = significantly different than lockout position p <0.001
Figure 1. Example of lifter in MTP position (left) compared to lockout (right)
Beckham et al. Isometric Strength of Powerlifters in Key Positions of the Conventional Deadlift 35
forces produced in the MTP position over the other three
positions may be explained by a number of things. First,
simple observation showed a marked difference in position
even between the MTP and the position of second greatest
force (lockout). The MTP position is rather upright, with the
knees bent. Powerlifters, in mimicking the deadlift, are in a
relatively straight legged position and somewhat bent over the
bar. Figure 1 shows an example of the differing position for
one of the athletes. The lockout position likely creates a
greater moment on the lower back and hips, which may limit
performance. The greater knee bend used in the MTP position
probably provides a force-production advantage, as the
powerful extensor forces of the quadriceps muscles are used to
a greater extent. The gluteus maximus may also be in a more
favorable position for resultant force production against the
bar, assuming a smaller hip moment, as was found in Brown &
and Escamilla et al.
Powerlifters in this study generated substantially different
amounts of force in each position. Changing mechanical
advantages probably contribute to the difference in forces, but
further research is needed to confirm this. Despite the
advantage of regular training in the deadlift-specific positions
(floor, knee and lockout), lifters still generated far more force
in the MTP. The MTP appears to represent the position of
greatest force output, even in lifters who train regularly in the
other positions. Lower force production capabilities in the
lower positions represent a limiting factor for deadlift
performance, thus an emphasis in training of the lower ranges
of motion of the deadlift may elicit greater gains.
1. Escamilla RF, Francisco AC, Fleisig GS, Barrentine SW, Welch CM,
Kayes AV, et al. A three-dimensional biomechanical analysis of sumo and
conventional style deadlifts. Med Sci Sports Exerc 2000; 32: 1265-75.
2. Brown EW, Abani K. Kinematics and kinetics of the dead lift in adolescent
power lifters. Med Sci Sports Exerc 1985; 17: 554-63.
3. Hales ME, Johnson BF, Johnson JT. Kinematic analysis of the powerlifting
style squat and the conventional deadlift during competition: is there a
cross-over effect between lifts? J Strength Cond Res 2009; 23: 2574-80.
4. McGuigan MR, Wilson BD. Biomechanical analysis of the deadlift. J
Strength Cond Res 1996; 10: 250-5.
5. Haff GG, Stone MH, O’Bryant H, Harman E, Dinan C, Johnson R, et al.
Force-time dependent characteristics of dynamic and isometric muscle
actions. J Strength Cond Res 1997; 11: 269-72.
6. Kraska JM, Ramsey MW, Haff GG, Fethke N, Sands WA, Stone ME, et al.
Relationship between strength characteristics and unweighted and
weighted vertical jump height. Int J Sports Physiol Perform 2009; 4: 461-
7. Juneja H, Verma SK, Khanna GL. Isometric strength and its relationship to
dynamic performance: a systematic review. J Exerc Sci Physio 2010; 6:
8. Challis JH. The appropriate scaling of weightlifting performance. J
Strength Cond Res 1999; 13: 367-71.
9. Hopkins W. A new view of statistics. 2002. Retrieved from
10. Skinner JW. Kinetic and kinematic analysis of the squat with and without
supportive equipment: Thesis, Appalachian State University; 2011.