USING SQUAT TESTING TO PREDICT TRAINING
LOADS FOR THE DEADLIFT, LUNGE, STEP-UP,
AND LEG EXTENSION EXERCISES
WILLIAM P. EBBEN, CHRISTINA R. FELDMANN, ANDREA DAYNE, DIANA MITSCHE,
LAUREN M. CHMIELEWSKI, PAUL ALEXANDER, AND KENNETH J. KNETGZER
Strength and Conditioning Research Laboratory, Department of Physical Therapy/Program in Exercise Science,
Marquette University, Milwaukee, Wisconsin
ABSTRACT
Ebben, WP, Feldmann, CR, Dayne, A, Mitsche, D, Chmielewski,
LM, Alexander, P, and Knetgzer, KJ. Using squat testing to
predict training loads for the deadlift, lunge, step-up, and leg
extension exercises (squat regression study). J Strength Cond
Res 22(6): 1947–1949, 2008—The purpose of this study was
to determine whether there is a linear relationship between the
squat and a variety of quadriceps resistance training exercises
for the purpose of creating prediction equations for the
determination of quadriceps exercise loads based on the squat
load. Six-repetition maximums (RMs) of the squat, as well as
four common resistance training exercises that activate the
quadriceps including the deadlift, lunge, step-up, and leg
extension, were determined for each subject. Subjects included
21 college students. Data were evaluated using linear
regression analysis to predict quadriceps exercise loads from
6RM squat data and were cross-validated with the prediction of
sum of squares statistic. Analysis of the data revealed that the
squat is a significant predictor of loads for the dead lift (R2 =
0.81, standard error of the estimate [SEE] = 12.50 kg), lunge
(R2 = 0.62, SEE = 12.57 kg), step-up (R2 = 0.71, SEE = 9.58
kg), and leg extension (R2 =0.67, SEE = 10.26 kg) exercises.
Based on the analysis of the data, the following 6RM prediction
equations were devised for each exercise: (a) deadlift load =
squat load (0.83) + 14.92 kg, (b) lunge load = squat load
(0.52) + 14.82 kg, (c) step-up load = squat load (0.50) + 3.32
kg, and (d) leg extension load = squat load (0.48) + 9.58 kg.
Results from testing core exercises such as the squat can provide
useful data for the assignment of loads for other exercises.
KEY WORDS strength, testing, load, program design
Address correspondence to William P. Ebben, webben70@hotmail.com.
22(6)/1947–1949
Journal of Strength and Conditioning Research
Ó 2008 National Strength and Conditioning Association
Copyright © N ational S trength and Conditioning A ssociation. Unauthorized reproduction of this article is prohibited.
INTRODUCTION
R
esistance training program design includes the
prescription of exercise load (1). Strength and
conditioning coaches typically use either trial and
error or repetition maximum (RM) testing to
prescribe loads for their athletes (5,6,7,11). Although training
loads are often based on testing data, it is impractical and
uncommon for practitioners to test numerous auxiliary
exercises (5,6,7,11). Therefore, regression analysis, and the
resultant prediction equations, offers a method of prescribing
exercise loads without having to test numerous auxiliary
exercises or rely on trial and error.
Previous studies have used regression analysis to predict
sprint times from resistance training data (2) and to assess the
accuracy of 1RM prediction equations based on RM bench
press testing (16). Studies have also used regression analysis
for prescribing resistance training loads. For example,
regression analysis has been used to predict free weight
squat loads based on the loads used with the angled leg press
(16), to predict the leg press (12), and to use the squat to
predict the Smith machine squat loads (3). Only one previous
study has demonstrated that resistance training core
exercises such as the squat can be used to predict a variety
of lower-body auxiliary exercises loads. This study deter-
mined that the squat is effective at predicting loads for
a variety of exercises that activate the hamstring muscle
group as the prime mover (8). As a result, prediction
equations were created to determine the loads for the
auxiliary exercises based on the squat load. Up to this point,
no previous research has determined whether resistance
training core exercises could be used to predict auxiliary
exercise loads thought to activate the quadriceps as the prime
mover. Therefore, the purpose of this study is to determine
whether there is a linear relationship between squat loads and
the loads used for a variety of lower-body resistance training
exercises thought to activate the quadriceps muscle group as
a prime mover. This study will also create prediction
equations for the determination of the exercise loads for
the deadlift, lunge, step-up, and leg extension, based on the
squat load.
VOLUME 22 | NUMBER 6 | NOVEMBER 2008 | 1947
Squat Regression Study

techniques (4). Subjects performed each exercise at maximal

volitional velocity and were equally encouraged for each
TABLE 1. Mean (6 SD) values for selected subject 6RM test during the orientation and testing sessions.
characteristics. Subjects performed the 6RM tests for the back squat,
Variable All subjects (N = 21) deadlift, lunge, step-up, and leg extension in randomized order,
with 4 minutes of recovery between each test exercise. All
Age (y) 20.86 6 1.85 subjects attained at least six repetitions of the 6RM loads. Six-
Age range (y) 18–26
repetition maximums were chosen because assistance exer-
Weight (kg) 73.98 6 10.89
Weight range (kg) 57.82–93.18 cises such as the leg extension are not usually performed at
Squat 6RM (kg) 98.26 6 30.00 fewer than six repetitions according to published descriptions
Squat 6RM range (kg) 47.73–150.00 of periodized programs (14). Squat 1RMs were not tested
Deadlift 6RM (kg) 96.37 6 27.71 because of the possibly large number of test sets required to
Lunge 6RM (kg) 66.05 6 19.08
reach their true 1RM and because testing needed to be
Step-up 6RM (kg) 52.22 6 17.75
Leg extension 6RM (kg) 56.28 6 17.42 consistent between the predictor and criterion variables. The
test exercises included common resistance training exercises
6RM = six-repetition maximum. that require the use of external loads that could be quantified.
The leg extension exercise was included to allow comparison
of open– and closed–kinetic chain exercises.

METHODS Statistical Analyses

Experimental Approach to the Problem
This study was designed to test the hypothesis that a linear
relationship exists between the squat 6RM and the 6RMs of
a variety of lower-body quadriceps exercises, such that
prediction equations could be created for the purpose of
prescribing a variety of exercise loads from squat testing data.
The criterion variables included the loads used during the
performance of the 6RM deadlift, lunge, step-up, and leg
extension exercises. The predictor variable was the 6RM
squat load.
Subjects
Twenty-one collegiate athletes and recreationally active
students who participate regularly in lower-body resistance
training volunteered for this study. Athlete subjects were in
their off-season training programs. Nonathlete subjects
participated in lower-body resistance training exercises at
least twice a week for a minimum of 2 months. All subjects
trained with a variety of closed–kinetic chain exercise
including the ones tested in this study. Table 1 describes
the subjects’ characteristics. All subjects provided informed
consent before the study. This study was approved by the
institution’s review board.
Procedures
At the beginning of the pretest orientation and testing
sessions, subjects participated in a general and dynamic
warm-up, consistent with the methods previously used in
related studies (10). Subjects were instructed to not
participate in resistance training for 48 hours before the
orientation session and for approximately 72 hours between
the orientation and testing sessions. The pretest orientation
session was designed to determine the 6RMs and to
familiarize the subjects with the exercise technique and
experimental procedures. All subjects were instructed,
performed, and were spotted with previously recommended
the TM
The data from the investigation are presented as mean 6 SD.
The statistical analyses were undertaken with SPSS 15.0 for
Windows (SPSS, Inc. Chicago, Ill) using linear regression
analysis to determine whether squat load was a significant
predictor of deadlift, lunge, step-up, and leg extension exercise
load, and to develop prediction equations for each of the
exercises. Assumptions for linearity of statistics were tested
and met. The prediction equations for each of the exercises
were cross-validated using the prediction of sum of squares
(PRESS) statistic, as has been previously recommended, and
according to methods described, in the literature (9,13).
RESULTS
Analysis of the data revealed that the squat was a significant
predictor of the load for the deadlift, lunge, step-up, and leg
extension, with R2 values ranging from 0.67 to 0.81. Based on
the analysis of the data, prediction equations were devised
and can be used to calculate training loads for each exercise
(Table 2). Results of the cross-validation procedure using the
PRESS statistic indicated that the predicted and actual loads
were similar (Table 2).
DISCUSSION
This is the first known study to use core exercise testing results
to create prediction equations for a variety of exercises
that activate the quadriceps as the prime mover. High R2
values suggest that the squat is a good predictor of the
quadriceps exercise loads, and the results indicate that the
squat accounts for more of the percentage of variance in the
quadriceps exercises than a previous study that used regression
analysis to predict squat load from leg press load, body mass,
and limb length (16). The standard error of the estimate values
in the present study indicate that the regression equations are
valid, despite a fairly heterogeneous sample. These results offer
a method of determining exercise loads that is more precise

1948 Journal of Strength and Conditioning Research

Copyright © N ational S trength and Conditioning A ssociation. Unauthorized reproduction of this article is prohibited.

TABLE 2. Six-repetition maximum prediction equations, regression values, and prediction of sum of squares (PRESS)
statistic cross-validation for each exercise (N = 21).
Exercise Equation
Deadlift Deadlift load = squat load (0.83) + 14.92 kg
Lunge Lunge = squat load (0.52) + 14.82 kg
Step-up Step-up = squat load (0.50) + 3.32 kg
Knee extension Leg extension = squat load (0.48) + 9.58 kg
SEE = standard error of the estimate.
than trial and error or subjective methods. The high R2 value
for the deadlift is thought to be attributable to the bio-
mechanical similarities between this exercise and the squat.
It is interesting to note that even open–kinetic chain exercises
such as the leg extension are predictable from biomechanically
dissimilar closed–kinetic chain exercises such as the squat. Not
surprisingly, previous research found slightly lower R2 values
when the squat was used to predict training loads for
hamstring-dominant lower-body exercises. These R2 values
ranged from 0.58 to 0.82 (8). Consequently, a variety of lower-
body exercises loads can be predicted from the squat load.
In the future, regression analysis may be useful for
predicting loads for a variety of other assistance exercises
based on core exercise test data. The results of this study are
most generalizable to exercises performed at loads that are
similar to those used in this study and with individuals whose
physical characteristics are similar to those of the subjects in
this study.
PRACTICAL APPLICATIONS
Exercise load assignment should be based on testing data as
opposed to guessing or trial and error. It is impractical and
inefficient for coaches or practitioners to test athletes in
numerous exercises to prescribe loads. As a result of this
study, load assignment for quadriceps exercises can be made
based on squat testing data, using the formulas provided.
ACKNOWLEDGMENTS
This study was funded by a Marquette University College of
Health Science Faculty Development Research Grant.
REFERENCES
1. Baechle, TR, Earle, RW, and Wathen, D. Resistance training. In:
Essentials of Strength Training and Conditioning. Baechle, TR, and
Earle, RW, eds. Champaign: Human Kinetics, 2000. pp. 395–426.
2. Blazvich, AJ and Jenkins, DG. Predicting sprint running times from
isokinetic and squat lift tests: a regression analysis. J Strength Cond
Res 12: 101–103, 1998.
3. Cottereman, ML, Darby, LA, and Skelly, WA. Comparison of
muscle force production using Smith machine and free weights
for bench press and squat exercises. J Strength Cond Res 19:
169–176, 2005.
Copyright © N ational S trength and Conditioning A ssociation. Unauthorized reproduction of this article is prohibited.
the TM
Journal of Strength and Conditioning Research | www.nsca-jscr.org
R2 SEE (kg) R2p SEEp(kg)
0.81 12.50 0.82 11.44
0.62 12.57 0.63 11.82
0.71 9.85 0.71 9.40
0.67 10.26 0.68 9.69
4. Earle, RW and Baechel, TR. Resistance training and spotting
techniques. In: Essentials of Strength Training and Conditioning.
TR. Baechle and RW. Earle, eds. Champaign: Human Kinetics. 2000.
pp. 343–389.
5. Ebben, WP and Blackard, DO. Strength and conditioning practices
of National Football League strength and conditioning coaches.
J Strength Cond Res 15: 48–58, 2001.
6. Ebben, WP, Carroll, R, and Simenz, C. Strength and conditioning
practices of National Hockey League strength and conditioning
coaches. J Strength Cond Res 18: 889–897, 2004.
7. Ebben, WP, Hintz, MJ, and Simenz, C. Strength and conditioning
practices of Major League Baseball strength and conditioning
coaches. J Strength Cond Res 19: 538–546, 2005.
8. Ebben, WP, Long, NJ, Pawlowski, ZD, Chmielewski, LM, Clewien, R,
and Jensen, RL. Using squat RM testing to determine hamstring
resistance training exercise loads. J Strength Cond Res (in press).
9. Holiday, D, Ballard, JE, and McKeown, BC. PRESS-related statistics:
regression tools for cross validation and case diagnosis. Med Sci Sports
Exerc 27: 612–620, 1995.
10. Jensen, RL and Ebben, WP. Hamstring EMG response during the
backsquat at different knee angles during concentric and eccentric
phases. In: Proceedings of the XVIII International Symposium of the
Society of Biomechanics in Sports. Hong Kong, Department of Sports
Science and Physical Education, The Chinese University of Hong
Kong, 2000. pp. 158–161.
11. Simenz, C, Dugan, C, and Ebben, WP. Strength and conditioning
practices of NBA strength and conditioning coaches. J Strength Cond
Res 19: 495–504, 2005.
12. Simpson, SR, Rozenek, R, Garhammer, J, and Lacourse, M.
Comparison of one repetition maximums between free weight and
Universal machine exercise. J Strength Cond Res 11: 103–106, 1997.
13. Sun, SS, Chumlea, WC, Heymsfield, SB, Lukaski, HC, Schoeller, D,
Friedl, K, Kuczmarki, RJ, Flegal, KM, Johnson, CL, and Hubbard, VS.
Development of bioelectrical impedance analysis prediction equa-
tions for body composition with the use of a multi-component
model for use in epidemiologic surveys. Am J Clin Nutr 77:
331–340, 2003.
14. Wathen, D, Baechle, TR, and Earle, R. Training variation:
periodization. In: Essentials of Strength Training and Conditioning.
Baechle, TR, and Earle, RW, eds. Champaign: Human Kinetics.
2000. pp. 513–527.
15. Whisenant, MJ, Panton, LB, East, WB, and Broeder, CE. Validation
of submaximal prediction equations for the 1 repetition maximum
bench press test on a group of collegiate football players. J Strength
Cond Res 17: 221–227, 2003.
16. Willardson, JM and Bressel, E. Predicting a 10 repetition maximum
for the free weight parallel squat using the 45° angle leg press.
J Strength Cond Res 18: 567–571, 2004.
VOLUME 22 | NUMBER 6 | NOVEMBER 2008 | 1949