Journal of Science and Medicine in Sport (2009) 12, 310—316

ORIGINAL PAPER

The reliability of the 1RM strength test for

untrained middle-aged individuals
Itamar Levinger a,∗, Craig Goodman a, David L. Hare b,
George Jerums c, Deidre Toia d, Steve Selig a

a Centre for Ageing, Rehabilitation, Exercise and Sport, School of Human Movement,
Recreation and Performance, Victoria University, Australia
b Department of Cardiology and University of Melbourne, Austin Health, Australia
c Department of Endocrinology and University of Melbourne, Austin Health, Australia
d Department of Cardiology, Austin Health, Australia

Received 18 April 2007 ; received in revised form 2 October 2007; accepted 18 October 2007

KEYWORDS Summary The one-repetition maximum (1RM) test is considered the gold standard
Reliablity; for assessing muscle strength in non-laboratory situations. Since most previous 1RM
Resistance training; reliability studies have been conducted with experienced young participants, it is
Familiarisation; unclear if acceptable test—retest reliability exists for untrained middle-aged indi-
Strength testing viduals. This study examined the reliability of the 1RM strength test of untrained
middle-aged individuals. Fifty-three untrained males (n = 25) and females (n = 28)
aged 51.2 ± 0.9 years participated in the study. Participants undertook the first 1RM
test (T1) 4—8 days after a familiarisation session with the same exercises. 1RM
was assessed for seven different exercises. Four to eight days after T1, partici-
pants underwent another identical 1RM test (T2). Ten weeks later, 27 participants
underwent a third test (T3). Intraclass correlation coefficients (ICC), typical error
as a coefficient of variation (TEcv), retest correlation, repeated measures ANOVA,
Bland—Altman plots, and estimation of 95% confidence limits were used to assess
reliability. A high ICC (ICC > 0.99) and high correlation (r > 0.9) were found for all
exercises. TEcv ranged from 2.2 to 10.1%. No significant change was found for six
of the seven exercises between T1 and T2. Leg press was slightly higher at T2 com-
pared to T1 (1.6 ± 0.6%, p = 0.02). No significant change was found between T2 and
T3 for any exercise. 1RM is a reliable method of evaluating the maximal strength
in untrained middle-aged individuals. It appears that 1RM-testing protocols that
include one familiarisation session and one testing session are sufficient for assessing
maximal strength in this population.
© 2007 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.

Introduction
∗
Corresponding author.
E-mail address: [Link]@[Link] Resistance training has been of great interest
(I. Levinger). to exercise scientists and health professionals

1440-2440/$ — see front matter © 2007 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/[Link].2007.10.007
The reliability of the 1RM strength test for untrained middle-aged individuals 311

who study both healthy individuals, and those Some investigators have suggested that older indi-
who suffer from chronic conditions including car- viduals should undergo between 8—9 sessions of
diovascular diseases,1—3 diabetes mellitus,4,5 and 1RM testing in order to increase the consistency
overweight/obese individuals.6 Since an increase of the 1RM measurements.23 Multiple familiarisa-
in muscle strength is the most common and tion sessions, however, may not be practical for
important benefit of resistance training,7,8 an training studies that examine strength changes for
accurate determination of muscle strength to several different resistance exercises as well as
properly evaluate the efficacy of the training is numerous other functional parameters. This would
essential.9 Laboratory-based methods for evalu- unnecessarily increase the time requirement of
ating maximal muscle strength include the use each participant and extend the duration of the
of isometric dynamometers10,11 and isokinetic study, both of which might increase the likelihood
dynamometers.12—14 These two methods, however, of participant dropout. As such, the purpose of this
usually require sophisticated laboratory equipment study was to examine the test—retest reliability
and personnel trained in their use. Furthermore, of the 1RM strength test of untrained middle-
these tests are not very specific for the types aged individuals following one familiarisation
of movement patterns commonly used in typical session.
fitness regimes. In contrast, the one-repetition
maximum (1RM) method, defined as the maximal
weight that can be lifted once with correct lift- Methods
ing technique, is comparatively simple and requires
relatively inexpensive non-laboratory equipment.9 Fifty-three untrained males (n = 25) and females
Moreover, because the 1RM test can be performed (n = 28) aged 51.2 ± 0.9 years volunteered to par-
using the same patterns as those undertaken by ticipate in the study (participants’ characteristics
the exercising individuals during their normal train- are shown in Table 1). Participants were included,
ing, it is increasingly gaining acceptance as the whether or not they had cardiovascular risk factors
gold standard for assessing muscle strength.8,9 Fur- such as overweight/obesity, hypertension, dyslipi-
thermore, previous studies have reported that the demia and hyperglycaemia. Participants were on a
1RM method to assess muscle strength is safe for range of medications including beta-blockers (two
healthy adults15,16 and also for patients with car- participants), calcium channel blockers (two), ACE
diovascular disease.17,18 The test—retest reliability inhibitors (four), diuretics (one), statins (two),
of the 1RM demonstrates high intraclass correlation metformin (one), and hormone replacement ther-
coefficients (ICC). However, as most 1RM reliabil- apy (six). They were only included if they had not
ity studies have been conducted with experienced been involved in regular aerobic physical activity
healthy young participants (age 18—30 years),19—21 in the previous 6 months or resistance training in
it is unclear whether this test—retest reliability the previous 5 years or more. Participants were
is applicable to untrained middle-aged individuals excluded from the study if they were involved in
who are increasingly the subject of exercise inter- vigorous regular exercise activity for more than
vention studies. It has been suggested that the 90 min a week or had documented heart disease
reliability of strength tests in older populations may such as chronic heart failure or coronary artery
be lower due to decreased muscle strength and disease. Each participant received an explanation
joint stability.12 To date, there is a lack of data on about the nature of the study and after a medical
the test—retest reliability of 1RM tests performed clearance signed an informed consent document.
by untrained middle-aged individuals for a range The study protocol was approved by the Human
of different resistance exercises. Most studies that Research Ethics Committees of both Victoria Uni-
have examined the test—retest reliability of maxi- versity and Austin Health.
mal strength in middle-aged and older populations Four to eight days prior to the first one-repetition
have used isokinetic dynamometers12,22,23 and not maximum test (T1), participants performed a
isoinertial-based (gym) machines. familiarisation session with the resistance training
A familiarisation process prior to 1RM strength equipment (Life Fitness and CalGym, Caloundra,
testing is essential for ensuring reliable test QLD, Australia). During the familiarisation session,
results24,25 and minimize learning effect or system- correct lifting and breathing technique were taught
atic bias.26 Furthermore, it has been shown that and practiced using submaximal and near-maximal
without a familiarisation process prior to strength loads. 1RM was defined as the heaviest weight a
testing, there is a significant increase in the expres- participant could lift once with a proper lifting
sion of muscle strength between two consecutive technique, without compensatory movements. 1RM
strength tests performed a few days apart.24,25 strength was assessed for seven different exercises
312 I. Levinger et al.

Table 1 Participants’ characteristics (mean ± S.E.M.)

All participants Subgroup
Gender (M/F) 25/28 13/14
Age (years) 51.2 ± 0.9 (40—69) 50.7 ± 1.3 (40—63)
Height (cm) 168.5 ± 1.3 (152.0—186.0) 168.1 ± 1.6 (155—185)
Weight (kg) 78.6 ± 2.3 (40—113.6) 78.5 ± 3.3 (40—112.4)
Waist (cm) 91.9 ± 1.9 (59.1—120.5) 90.9 ± 0.9 (59.1—120.5)
BMI (kg m−2 ) 27.5 ± 0.7 (16.7—39.22) 27.5 ± 2.6 (16.7—36.3)
Note: BMI, body mass index.

comprising in order, chest press, leg press, lat- Also, we report the lower and upper confidence
eral pull-down, triceps pushdown, knee extension, limits (95%) of the above measurements. Changes
seated row and biceps curl. Exercises involving in the means between tests, which examines ran-
large muscle groups were performed first followed dom change and systematic bias, were assessed
by those involving small muscle groups. In order by repeated-measures ANOVA. Log transformation
to facilitate the recovery and reduce the effect (multiplied by 100 to maintain precision) was per-
of fatigue, exercises were alternated between the formed prior to the data analyses as data were
upper and lower body. The tests commenced after not normally distributed. Data distribution was
a light warm-up (3-min walking at self-selected examined by normal probability plots and the
speeds on a treadmill). The maximal strength test Kolmogorov—Smirnov test because heteroscedas-
protocol included one set of 10 repetitions at a rel- tic errors are the norm in studies in the area
atively light load that served as a specific warm-up, of exercise science.27,29 Log transformation was
followed by a gradual increase in load until 1RM was performed using a spreadsheet designed for this
achieved. The rate of the gradual increase in load purpose.26,30 Bland—Altman plots were used to dis-
was dependent on the participant’s self-perceived play visual representations of the errors against
capacity, and it ranged from 1 to 10 kg for biceps true values by plotting the difference between
curl and triceps pushdown, up to 100 kg for leg press T2 and T1 against the mean of T2 and T1.31
and between 1 and 20 kg for the rest of the exer- Data were analysed using SPSS (Version 15). Data
cises. The 1RM was achieved within 3—6 attempts. are presented as means ± standard error of mea-
The rest period between attempts was 1 min, and surements (SEM), unless otherwise stated. All
between each specific exercise, volunteers recov- statistical analyses were conducted at the 95% level
ered for 2 min. Four to eight days after the T1, of significance.
participants underwent another 1RM test (T2) and,
for a subgroup of 27 participants, again after 10
weeks (T3). T2 and T3 test protocols and the test Results
conditions were identical to test T1,21 and the same
investigators conducted the testing on each occa- No significant injuries occurred during the study,
sion. The 27 participants from the subgroup were except for mild muscle soreness that is common
asked not to change their activity level and this was with unaccustomed exercise for untrained individ-
monitored. uals. Some participants reported this mild soreness
Reliability in this study was analysed according 1—3 days after some of the tests.
to the recommendation of Hopkins26 and Atkin- Raw data of T1 and T2 for all seven exercises
son and Nevil.27 Intraclass correlation coefficients (53 participants) are presented in Table 2. Also
were used as they are a common measurement reported are the ICC and the change in mean, TEcv
of reliability that enables comparisons with other (%) and the correlation between the two tests. A
studies.27 The ICC method was based on a repeat high ICC (>0.97) and high correlation (r > 0.9) were
measurement of maximal strength (single value) (2, found for all exercises and for the total strength
1) and the same investigator conducted the tests.28 (sum of all seven exercises). TEcv (%) indicated
Retest correlation was measured by Pearson cor- typical errors range of 2.2—7.2%. In addition, no
relation coefficient. Within-subject variation was significant systematic bias was found between total
reported as typical error, and was expressed as weight lifted at T1 and T2, nor for six of the
a coefficient of variation (TEcv). This allows for individual exercises including chest press, triceps
comparisons between different studies, and is usu- pushdown, seated row, leg extension, biceps curl
ally less affected by machine or operator error.26 and lateral pull (Table 2). There was a significant
The reliability of the 1RM strength test for untrained middle-aged individuals 313

Table 2 Muscle strength and the reliability for the 53 participants who underwent T1 and T2
Exercise T1 (kg ± S.E.) T2 (kg ± S.E.) ICC Reliability after log-transformation

 mean (%) (95% CI) TE as a CV (%) (95% CI) Retest r

Chest P 52.5 ± 3.5 52.5 ± 3.6 0.99 −0.7 (−3.1 to 1.7) 6.5 (5.4 to 8.1) 0.984
Leg P 195 ± 11.2 198.8 ± 11.4 0.99 1.6 (0.3 to 2.9)* 3.3 (2.8 to 4.1) 0.993
Lat P 50.5 ± 2.3 50.0 ± 2.3 0.99 −1.1 (−2.3 to 0.2) 3.4 (2.8 to 4.2) 0.990
Tri P 27.4 ± 1.4 27.7 ± 1.4 0.98 0.9 (−1.1 to 2.9) 5.3 (4.4 to 6.6) 0.981
Knee E 46.9 ± 2.1 48.0 ± 2.2 0.97 1.7 (−0.7 to 4.2) 6.0 (5.0 to 7.6) 0.969
Seated R 50.8 ± 2.4 50.3 ± 2.4 0.99 −1.1 (−2.4 to 0.2) 3.4 (2.8 ± 4.2) 0.992
Biceps C 18.8 ± 1.1 19.1 ± 1.2 0.98 1.1 (−1.6 to 3.9) 7.2 (6.0 to 9.0) 0.979
All 434.3 ± 21.0 438.2 ± 21.5 0.99 0.7 (−0.2 to 1.6) 2.2 (1.9 to 2.8) 0.997
Note: ICC, intraclass correlation coefficients; TE, typical error; CV, coefficient of variation; Chest P, chest press; Leg P, leg press;
Lat P, lateral pull-down; Tri P, triceps push down; Knee E, knee extension; Seated R, seated row; Biceps C, biceps curl; Total, total
weight lifted.
* p < 0.05.

Figure 1 (A) Bland—Altman plot for all 53 participants. (B) Bland—Altman plot for 27 subgroup of participants that
also were tests at T3. Solid line represents mean difference between two tests. Dashed lines represent  difference
between two tests ± 2 S.D. For both panels data point represents comparison for the total strength (sum of all seven
exercises).

but small difference in leg press 1RM with T2 being that the 1RM test, after one familiarisation ses-
slightly higher than T1 (1.6 ± 0.6%, p = 0.02). The sion, using standard resistance training equipment,
Bland—Altman plot within cases for T1 and T2 for is a reliable and simple tool for assessing maxi-
total strength demonstrated also that most results mal strength for untrained middle-aged individuals
for individual participants are within 2 standard across a wide array of resistance exercises.
deviation (Fig. 1A.). To our knowledge, this study is the first to exam-
The reliability data of the 27 individuals in the ine the reliability of the 1RM test in a wide range
subgroup who performed T1, T2 and T3 are pre- of isoinertial resistance exercises, compared to
sented in Table 3. ICC was >0.96 between T2 and only one or two exercises in previous studies. The
T1 and T3 and T2. TEcv (%) range was 1.9—10.1%. In present study had a relatively large sample size
addition, the retest correlation was >0.9. No signifi- (n = 53) compared to other studies that examined
cant systematic bias was found between any of the the reliability of the 1RM test.21,23,32 The present
exercises. Most data points on the Bland—Altman study demonstrated high ICC across all exercises
plot for total strength were within 2 standard devi- examined with inexperienced middle-aged individ-
ation for T2 versus T3 (Fig. 1B). uals. A high ICC was observed both between T1 and
T2 and T2 and T3. These results are similar to the
reported test—retest reliability of the 1RM test for
Discussion experienced resistance-trained young individuals.
The ICC of 1RM testing is specific to the exercise
Accurate evaluations of muscle strength are impor- and ranges between 0.64 and 0.99 for exercises
tant to prescribe safe and effective resistance such as leg press,20,21,32 bench press,20,21,33 lat-
training intensities and to evaluate the efficacy of eral pull-down32 and leg extension.32 However, it
training. The main finding of the current study is has been suggested that ICC may overestimate
314 I. Levinger et al.

reliability, especially when the sample data are

Retest r

Note: ICC, intraclass correlation coefficients; TE, typical error; CV, coefficient of variation; Chest P, chest press; Leg P, leg press; Lat P, lateral pulldown; Tri P, triceps push down; Knee
heterogeneous.27 In this study we found high retest

0.972
0.993
0.996
0.976
0.979
0.995
0.978
0.958
correlation (r > 0.9) for middle-aged individuals.
Previously, Frontera et al.24 reported a retest cor-
relation in the range of 0.7—0.8, using isokinetic
dynamometer muscle strength testing in elderly

TE as a CV (%) (95% CI)

individuals (mean age = 60 years, range 45—78). It is
Reliability after log-transformation (T3 vs. T2)

14.1)

10.7)
12.1)
important to note that the test protocol of Frontera

5.4)
3.3)
8.9)
7.6)
3.9)
to et al.24 did not include a familiarisation session,
to
to
to
to
to
to
to
and as such, it is possible that one familiarisation
(7.9
(3.0
(1.9
(5.0
(4.2
(2.2
(6.0
(6.8
session (as used in the current study) improved the
correlation (reliability) between the two tests.
10.1
3.9
2.4
6.5
5.4
2.8
7.7
8.7
Typical error (within-subject variation) is an
important measure of reliability in the field of
exercise and sports science, and it reflects the

E, knee extension; Seated R, seated row; Biceps C, biceps curl; Total, total weight lifted. Data were analysed after log transformation.
change in scores between tests.26 Hopkins (2000)
 mean (%) (95% CI)

−0.1)

recommended reporting of the typical error as

3.9)
3.2)
1.0)
4.4)
0.7)
1.1)

6.7)

coefficient of variation (TEcv) as it allows direct

to
to
to
to
to
to
to
to

comparisons between different studies, and it is

(−6.7
(−1.1
(−1.6
(−2.6
(−6.7
(−2.0
(−8.1
(−2.8

also less affected by different machines, analy-

sers or diverse cohorts of participants. The lower
−1.6
1.0
−0.3
0.8
−3.7
−0.5
−4.2
1.8

the TEcv value, the higher the reliability (less

variation).34 It is difficult to asses if the TEcv in this
study indicates high, good or low reliability, as we
found only one other study that reported this value
ICC (T3—T2)

for maximal strength testing.34 Symons et al.34

examined the reliability of concentric, isometric
and eccentric isokinetic knee extensor strength in
Muscle strength and reliability for the 27 participants who underwent T3

0.97
0.99
0.99
0.98
0.97
0.99
0.97
0.96

older women. They reported that the TEcv ranged

from 7 to 11% for the knee extensors. The authors
concluded that the interpretation depends on the
precision required but that this range was adequate
458.07 ± 32.7
T3 (kg ± S.E.)

209.7 ± 16.6
55.0 ± 5.9

52.6 ± 3.7
28.7 ± 2.2
47.3 ± 3.5
52.7 ± 4.0
19.3 ± 1.7

for assessing muscle strength before training. In

the current study, we found lower values between
T1 and T2 (2.2—6.5%) and between T2 and T3
(2.4—10.1%), compared to Symons et al.34 The dif-
ference in the TEcv range may be due to the lack of
a familiarisation session in their protocol testing.
Changes in the mean (systematic bias) also
T2 (kg ± S.E.)

203.4 ± 15.8

446.2 ± 32.4
54.5 ± 5.7

52.3 ± 3.6
28.4 ± 2.2
48.6 ± 3.2
52.4 ± 3.8
19.6 ± 1.7

indicate consistency between tests. It has been

assumed that an increase in muscle strength
between two tests can be the result of a learn-
ing effect and as such, several familiarisation trials
may be necessary before applying an exercise train-
ing intervention.26 In the current study, there were
no significant changes in the maximal weight lifted
T1 (kg ± S.E.)

201.1 ± 15.5

442.1 ± 31.9
53.9 ± 5.6

52.5 ± 3.6
28.0 ± 2.1
46.8 ± 3.1
53.0 ± 3.8
19.4 ± 1.7

from T1 to T2 for six of seven exercises. The only

difference was observed in leg press exercise with
the T2 being slightly higher than T1. However, this
difference between leg press strength at T1 and
T2 was small (representing a difference of only
1.9% or ∼3 kg) compared to the expected change
in strength following resistance training of least
Seated R
Biceps C
Exercise
Table 3

Chest P

15%.9 Further evidence suggesting that minimal or

Knee E
Leg P
Lat P

Total
Tri P

no learning effects took place in the present study

may be seen for the participants that performed
The reliability of the 1RM strength test for untrained middle-aged individuals 315

a third 1RM test (T3). No significant improvement process may lead to an increase in the dropout
was observed for all exercises from T2 to T3 (see rates of participants, particularly those with low or
Table 3). marginal levels of motivation40 who would benefit
Our study also demonstrates that similar to the most from clinical intervention studies.
healthy young individuals,32,33 one familiarisa- Practical implications
tion session may be sufficient before assessing
maximal strength in an untrained middle-aged • The 1RM method is a reliable and simple
population. Previously we have shown that one method to evaluate maximal strength in wide
familiarisation session is adequate for assess- array of resistance exercises in untrained
ing maximal strength (isokinetic dynamometer) middle-aged individuals.
in elderly clinical populations.25 However, Ploutz- • 1RM-testing protocols that include one famil-
Snyder and Giamis23 reported that older individuals iarisation session and one testing session
should undergo between eight and nine sessions of separated by 4—8 days are sufficient for
1RM testing (isokinetic dynamometer) in order to assessing maximal strength in this untrained
increase the consistency of the 1RM measurement. population.
It is important to note that the Ploutz-Snyder and
Giamis23 protocol included a 1RM test every 48 h
over 2—3 weeks. This type of familiarisation pro-
tocol may actually induce a training effect due to
neural35 or even musculoskeletal adaptations. In
addition repeated maximal strength testing over
several days is likely to increase the risk of over- References
training,36 that could result in blunted adaptations
1. Levinger I, Roger B, Cody DV, et al. Resistance training for
once the training program finally begins. It is impor- chronic heart failure patients on beta blocker medications.
tant to acknowledge that in our study 1RM tests Int J Cardiol 2005;102:493—9.
were performed on resistance machines that limit 2. Brochu M, Savage P, Lee M, et al. Effects of resistance
the movements to the desired direction. As such, training on physical function in older disabled women with
it is possible that more complex movements with coronary heart disease. J Appl Physiol 2002;92:672—8.
3. Hare DL, Ryan TM, Selig SE, et al. Resistance exercise train-
‘‘free weights’’ may require more familiarisation ing increases muscle strength, endurance, and blood flow in
sessions. patients with chronic heart failure. Am J Cardiol 1999;83,
Similar to Ploutz-Snyder and Giamis,23 some 1RM 1674—1677, A1677.
test guidelines recommend using multiple familiari- 4. Castaneda C, Layne JE, Munoz-Orians L, et al. A randomized
sation sessions before assessing maximal strength in controlled trial of resistance exercise training to improve
glycemic control in older adults with type 2 diabetes. Dia-
order to avoid an improvement in muscle strength betes Care 2002;25:2335—41.
due to improved motor coordination or other neu- 5. Cauza E, Hanusch-Enserer U, Strasser B, et al. The rel-
ral adaptations.9 This may be valid for investigators ative benefits of endurance and strength training on
who wish to examine increases in muscle strength the metabolic factors and muscle function of people
following resistance training due solely to mus- with type 2 diabetes mellitus. Arch Phys Med Rehabil
2005;86:1527—33.
cle hypertrophy. However, to our knowledge, most 6. Joseph LJO, Trappe TA, Farrell PA, et al. Short-term mod-
trials with untrained or clinical populations use erate weight loss and resistance training do not affect
resistance training as a mode of clinical or ther- insulin-stimulated glucose disposal in postmenopausal
apeutic intervention.37 As such, changes in motor women. Diabetes Care 2001;24:1863—9.
coordination (i.e. inter-muscular co-ordination) 7. Hurley BF, Roth SM. Strength training in the elderly:
effects on risk factors for age-related diseases. Sports Med
and other neural adaptations should be considered 2000;30:249—68.
as part of the benefits of the training.38,39 In other 8. Fleck SJ, Kraemer WJ. Designing resistance training pro-
words, if the aim of the study is to examine the grams. Champaign, IL: Human Kinetics; 2004.
early neural adaptations as a result of resistance 9. Kraemer WJ, Ratamess NA, Fry AC, et al. Strength train-
training, then muscle strength should be assessed ing: development and evaluation of methodology. In: Maud
PJ, Foster C, editors. Physiological assessment of human
before chronic exposure to resistance exercise that fitness. Champaign, IL: Human Kinetics; 2006. p. 119—50.
occurs with long familiarisation processes. 10. Sagiv M, Hanson P, Besozzi M, et al. Left ventricu-
Furthermore, in training studies with clinical lar responses to upright isometric handgrip and deadlift
populations, strength testing is usually not the in men with coronary artery disease. Am J Cardiol
only outcome measure with other physiological, 1985;55:1298—302.
11. Elkayam U, Roth A, Weber L, et al. Isometric exer-
biochemical, functional and psychological param- cise in patients with chronic advanced heart failure:
eters also being evaluated. Due to the complexity hemodynamic and neurohumoral evaluation. Circulation
of these types of studies, a long familiarisation 1985;72:975—81.
316 I. Levinger et al.

12. Ordway NR, Hand N, Briggs G, et al. Reliability of knee 25. Selig SE, Carey MF, Menzies DG, et al. Reliability of isoki-
and ankle strength measures in an older adult population. netic strength and aerobic power testing for patients with
J Strength Cond Res 2006;20:82—7. chronic heart failure. J Cardiopul Rehab 2002;22:282—9.
13. Zachwieja JJ, Toffolo G, Cobelli C, et al. Resistance 26. Hopkins WG. Measures of reliability in sports medicine and
exercise and growth hormone administration in older science. Sports Med 2000;30:1—15.
men: effects on insulin sensitivity and secretion dur- 27. Atkinson G, Nevill AM. Statistical methods for assessing
ing a stable-label intravenous glucose tolerance test. measurement error (reliability) in variables relevant to
Metabolism 1996;45:254—60. sports medicine. Sports Med 1998;26:217—38.
14. Selig SE, Carey MF, Menzies DG, et al. Moderate-intensity 28. Portney LG, Watkins III JB. Foundations of clinical research:
resistance exercise training in patients with chronic heart applications to practice. Upper Saddle River, NJ: Prentice
failure improves strength, endurance, heart rate variability, Hall Health; 2000.
and forearm blood flow. J Card Fail 2004;10:21—9. 29. Nevill AM, Atkinson G. Assessing agreement between mea-
15. Shaw CE, MaCully KK, Posner JD. Injuries during the one surements recorded on a ratio scale in sports medicine and
repetition maximum assessment in the elderly. J Cardiopul sports science. Br J Sports Med 1997;31:314—8.
Rehab 1995;15:283—7. 30. Hopkins WG. Reliability: calculation and more. Available
16. Gordon NF, Kohl 3rd HW, Pollock ML, et al. Cardiovascular from: [Link]
safety of maximal strength testing in healthy adults. Am J (Accessed 2007); 2007.
Cardiol 1995;76:851—3. 31. Bland JM, Altman DG. Statistical methods for assessing
17. Ghilarducci LEC, Holly RG, Amsterdam EA. Effects of high agreement between two methods of clinical measurement.
resistance training in coronary artery disease. Am J Cardiol Lancet 1986;1:307—10.
1989;64:866—70. 32. Patterson P, Sharman J, Hitzelberger L, et al. Test-retest
18. Featherstone JF, Holly RG, Amsterdam EA. Physiologic reliability of selected LifeCircuite machines. J Strength
responses to weight lifting in coronary artery disease. Am Cond Res 1996;10:246—9.
J Cardiol 1993;71:287—92. 33. Giorgi A, Wilson GJ, Weatherby RP, et al. Functional iso-
19. McBride JM, Triplett-McBride T, Davie A, et al. The effect metric weight training: its effects on the development of
of heavy- vs. light-load jump squats on the develop- muscular function and the endocrine system over an 8-week
ment of strength, power, and speed. J Strength Cond Res training period. J Strength Cond Res 1998;12:18—25.
2002;16:75—82. 34. Symons TB, Vandervoort AA, Rice CL, et al. Reliability
20. Rhea MR, Ball SD, Phillips WT, et al. A comparison of lin- of isokinetic and isometric knee-extensor force in older
ear and daily undulating periodized programs with equated women. J Aging Phys Act 2004;12:525—37.
volume and intensity for strength. J Strength Cond Res 35. Sale DG. Neural adaptation to resistance training. Med Sci
2002;16:250—5. Sports Exerc 1988;20:S135—45.
21. Kraemer WJ, Ratamess N, Fry AC, et al. Influence of resis- 36. Fry AC, Kraemer WJ, van Borselen F, et al. Performance
tance training volume and periodization on physiological decrements with high-intensity resistance exercise over-
and performance adaptations in collegiate women tennis training. Med Sci Sports Exerc 1994;26:1165—73.
players. Am J Sports Med 2000;28:626—33. 37. Barry BK, Carson RG. The consequences of resistance train-
22. Heinonen A, Sievanen H, Viitasalo J, et al. Reproducibility ing for movement control in older adults. J Gerontol Bio Sci
of computer measurement of maximal isometric strength Med Sci 2004;59:730—54.
and electromyography in sedentary middle-aged women. 38. Carroll TJ, Riek S, Carson RG. Neural adaptations to resis-
Eur J Appl Physiol 1994;68:310—4. tance training: implications for movement control. Sports
23. Ploutz-Snyder LL, Giamis EL. Orientation and familiariza- Med 2001;31:829—40.
tion to 1RM strength testing in old and young women. J 39. Griffin L, Cafarelli E. Resistance training: cortical,
Strength Cond Res 2001;15:519—23. spinal, and motor unit adaptations. Can J Appl Physiol
24. Frontera WR, Hughes VA, Dallal GE, et al. Reliability of 2005;30:328—40.
isokinetic muscle strength testing in 45- to 78-year-old 40. Rejeski WJ, Kenney EA. Fitness motivation: preventing
men and women. Arch Phys Med Rehabil 1993;74:1181— participant dropout. Champaign, IL: Life Enhancement
5. Publications; 1988.

Available online at [Link]