Content uploaded by Mark D. Peterson
Author content
All content in this area was uploaded by Mark D. Peterson on Oct 02, 2018
Content may be subject to copyright.
950
Journal of Strength and Conditioning Research, 2005, 19(4), 950–958
q2005 National Strength & Conditioning Association
A
PPLICATIONS OF THE
D
OSE
-R
ESPONSE FOR
M
USCULAR
S
TRENGTH
D
EVELOPMENT
:AR
EVIEW OF
M
ETA
-A
NALYTIC
E
FFICACY AND
R
ELIABILITY FOR
D
ESIGNING
T
RAINING
P
RESCRIPTION
M
ARK
D. P
ETERSON
,
1
M
ATTHEW
R. R
HEA
,
2
AND
B
RENT
A. A
LVAR
1
1
Department of Exercise and Wellness, Arizona State University, Tempe, Arizona 85287;
2
Department of Physical
Education, Southern Utah University, Cedar City, Utah 84720.
A
BSTRACT
.Peterson, M.D., M.R. Rhea, and B.A. Alvar. Appli-
cations of the dose-response for muscular strength development:
A review of meta-analytic efficacy and reliability for designing
training prescription. J. Strength Cond. Res. 19(4):950–958.
2005.—There has been a proliferation in recent scholarly dis-
cussion regarding the scientific validity of single vs. multiple
sets of resistance training (dose) to optimize muscular strength
development (response). Recent meta-analytical research indi-
cates that there exist distinct muscular adaptations, and dose-
response relationships, that correspond to certain populations.
It seems that training status influences the requisite doses as
well as the potential magnitude of response. Specifically, for in-
dividuals seeking to experience muscular strength development
beyond that of general health, an increase in resistance-training
dosage must accompany increases in training experience. The
purpose of this document is to analyze and apply the findings of
2 meta-analytical investigations that identified dose-response
relationships for 3 populations: previously untrained, recreation-
ally trained, and athlete; and thereby reveal distinct, quantified,
dose-response trends for each population segment. Two meta-
analytical investigations, consisting of 177 studies and 1,803 ef-
fect sizes (ES) were examined to extract the dose-response con-
tinuums for intensity, frequency, volume of training, and the
resultant strength increases, specific to each population. ES
data demonstrate unique dose-response relationships per popu-
lation. For untrained individuals, maximal strength gains are
elicited at a mean training intensity of 60% of 1 repetition max-
imum (1RM), 3 days per week, and with a mean training volume
of 4 sets per muscle group. Recreationally trained nonathletes
exhibit maximal strength gains with a mean training intensity
of 80% of 1RM, 2 days per week, and a mean volume of 4 sets.
For athlete populations, maximal strength gains are elicited at
a mean training intensity of 85% of 1RM, 2 days per week, and
with a mean training volume of 8 sets per muscle group. These
meta-analyses demonstrate that the effort-to-benefit ratio is dif-
ferent for untrained, recreationally trained, and athlete popu-
lations; thus, emphasizing the necessity of appropriate exercise
prescription to optimize training effect. Exercise professionals
may apply these dose-response trends to prescribe appropriate,
goal-oriented training programs.
K
EY
W
ORDS
. progression model, resistance exercise, perfor-
mance enhancement
I
NTRODUCTION
Current-day trends have cultivated a steady hi-
erarchical advancement in the professional dis-
ciplines of applied exercise science and preven-
tative health. The magnitude of available prac-
titioners, personal trainers, and sport conditioning coach-
es is increasing not only in sheer number, but also in
depth, as subspecialties within these professions emerge.
Unfortunately, this progression is not controlled, because
our field and the public fails to hold fitness professionals
to rigorous standards of knowledge, practice, and/or qual-
ification. The ensuing spread of tangential information
leads to confusion and discrepancy of fundamental prin-
ciples, as well as innovative scientific findings. Equipping
students of the exercise sciences with the appropriate
tools to convey sound exercise prescription and applica-
tion in the professional setting is the foremost imperative
step. However, by promoting the distribution of research
findings to the general exercising public, we can begin to
control the integrity and uniformity of this available in-
formation, and facilitate its propagation.
Moreover, the other end of the applied-fitness spec-
trum, which comprises the leaders of sport performance
enhancement, is highly specialized, but is also devoid of
thorough standardization. The conventional expectation
that professional and elite sport will continue to steadily
progress in grandeur encourages the ongoing conception
and refinement of training technologies and methodolo-
gies. This has forced performance-enhancement special-
ists, coaches, and athletes to build on established training
principles, and distinguish auxiliary components, often
without the endorsement of scientific investigation. Con-
sequently, athletic performance enhancement has devel-
oped into a volatile, yet especially vital institution and
determinant for today’s sport teams and individual ath-
letes.
Ostensibly, a positive side effect of growth in these
professions, and subsequent acknowledgment among the
public, is perpetual evolution of exercise-prescription
methods and implementation strategies. The design of an
effective training program is a complex process, however,
involving the application and synergism of established
scientific principles, progressive research findings, vet-
eran and modern practices, and professional knowledge
to accommodate individual situations, needs, and goals.
Whether the program is for a recreational exercise con-
sumer seeking improvements in muscular health and fit-
ness, or a professional athlete working for an advantage
on the field of play, ethics demand that sound exercise
prescriptions be used. To raise the standards, the Nation-
al Strength and Conditioning Association has worked to
bridge the gap between science and practice by providing
disciplinarians and professionals with research-based in-
formation that has applicable relevance. In support, the
A
PPLICATIONS OF THE
D
OSE
-R
ESPONSE
951
purpose of this paper is to discuss the findings of several
recent syntheses of strength development research and to
discuss the practical applications of these findings.
AD
OSE
-R
ESPONSE FOR
M
USCULAR
S
TRENGTH
:
U
NDERSTANDING THE
D
EBATE
A topic of scientific and professional significance that
both the American College of Sports Medicine and Na-
tional Strength and Conditioning Association have ac-
cepted is that of the dose-response for muscular strength
training and development (19–21). Historically, a quan-
tifiable relationship between the volume, intensity, and/
or frequency of training, and muscular strength improve-
ments, has been elusive and controversial. Although
much research has examined strength increases accom-
panying training interventions, most have examined only
1 or 2 training programs, providing only glimpses of a
dose-response relationship. Of the various prescription
components, volume of strength training has undoubtedly
received the most research attention (i.e., single vs. mul-
tiple sets). For many years, personal opinion and the ac-
counts of several unscientific literature reviews were the
primary sources of evidence to support training philoso-
phies.
Particularly, several narrative reviews (4, 9, 10) were
completed in the 1990s that examined the results of a
small number of studies using single and multiple-set
training programs. This type of review, which may be per-
ceived as more of an art than science, often relies heavily
on personal decisions of the reviewer, and contains nu-
merous areas in which bias may persuade the results.
Ultimately, such reviews have proven to be of little value
in resolving debate or confusion (3, 5). Not in contrast,
some of the conclusions and assumptions from the nar-
rative reviews regarding volume of training include: (a)
‘‘The opinion that multiple-set protocols are better than
a single set of an exercise is not supported by the consen-
sus of scientific evidence’’ (4); (b) ‘‘Multiple set training
has not been shown to provide additional benefits in the
adult fitness setting’’ (9); and (c) ‘‘The research results
show very convincingly that 1 set is generally as effective
as 3 sets’’ (9).
These conclusions, which are in accordance with the
philosophies of many low-volume advocates, presume a
universal relationship between volume of training and re-
sultant strength development, such that large initial im-
provements in muscular strength are achieved from the
first set of exercise, and very little, if any additional im-
provements are acquired with successive sets. In essence,
this view suggests that the human body is not capable of
hierarchical adaptation to greater stresses, beyond 1 set
of resistance training performed a few times per week.
Interestingly, in 2002, the American College of Sports
Medicine issued a position stand which highlighted the
topic of ‘‘Progression Models’’ for individuals seeking mus-
cular development beyond that of general health and fit-
ness (14). This position stand was an obligatory re-rec-
ommendation of guidelines for resistance-training pre-
scription. In 1998, the ACSM originally issued the sug-
gestion (1) that a single set of 8–10 exercises would result
in similar strength improvements as following a training
program of higher volume.
Even at that time, such conclusions drastically op-
posed training prescription practices, and seemed highly
speculative to many experienced members of the sport-
conditioning field (3). Furthermore, despite the more re-
cent comprehensive position stand on progressive train-
ing to accommodate and facilitate higher levels of mus-
cular fitness, the debate continues. In June, 2004, Car-
pinelli et al. released another narrative review, ‘‘A critical
analysis of the ACSM position stand on resistance train-
ing: Insufficient evidence to support recommended train-
ing protocols’’ (6). The purpose of this subsequent ‘‘objec-
tive’’ follow-up was to critically analyze the contents and
supportive rationale behind the 2002 ACSM Progression
Model. En route to accomplishing this task, the investi-
gators seemingly rebut each and every one of the defining
positions set forth by the ACSM. Specifically, regarding
training prescriptions for muscular strength, the authors
claim a copious failure of the cited evidence to support (a)
‘‘The superiority of free weights or machines for devel-
oping muscular strength, hypertrophy, power, or endur-
ance’’ (p. 5); (b) ‘‘The superiority of any specific repetition
duration for developing muscular strength, hypertrophy,
power, or endurance’’ (p. 9); (c) ‘‘The claim in the Position
Stand that specific ranges of repetitions produce specific
outcomes ...’’(p.11); (d) ‘‘. . . the superiority of multiple-
set training, while strongly supporting the efficacy of sin-
gle sets’’ (p. 18); (e) ‘‘The claim in the Position Stand that
the rest time between sets and exercise is dependent on
the specific goals of a particular exercise, that shorter rest
periods decrease the rate of strength gains, and that mul-
tiple-joint exercises require longer rest periods than sin-
gle-joint exercises on machines’’ (p. 19); and (f) ‘‘that the
planned manipulation of the program variables in ad-
vanced trainees can eliminate natural training plateaus
and enable higher levels of muscular strength, hypertro-
phy, power, and local muscular endurance’’ (p. 35).
Throughout the article, there is a preponderance of
the familiar biased dialog that often contaminates nar-
rative and critical reviews, confirming once again, the ob-
vious downfall of this clumsy (7) style of research. Clear-
ly, a more reliable, detailed, quantitative, and objective
method for reviewing very large numbers of research
studies is needed to shed brighter light on the dose-re-
sponse for strength, as well as other fitness endeavors.
M
ETA
-A
NALYTICAL
I
NVESTIGATION
:
Q
UANTIFYING THE
D
OSE
-R
ESPONSE
A powerful investigative method that researchers in sport
conditioning rarely use, despite proven efficacy in other
disciplines, is the meta-analysis. Nevertheless, this sta-
tistical review of the literature may potentially become a
principal tool to bridge the gap between the science and
practice of exercise prescription. The meta-analysis has
been used to synthesize treatment effects/effect size (ES)
data in many disciplines, including the medical, psycho-
logical, sociological, and other behavioral fields, to com-
bine the results of numerous similar studies, under sim-
ilar circumstances, and with similar participants, into 1
quantitative analysis. Furthermore, the meta-analysis al-
lows for an objective assessment of the magnitude of ef-
fectiveness for a given treatment or intervention through
established, meticulous statistical guidelines. This tech-
nique has served as a valuable research tool, facilitating
the observance of dose-response trends, wherein a contin-
uum of variable doses elicits a continuum of quantified
responses.
952 P
ETERSON
,R
HEA
,
AND
A
LVAR
F
IGURE
1. Intensity of training: the average percentage of 1
repetition maximum used throughout the training program.
F
IGURE
2. Frequency of training: the number of days per
week that participants trained a muscle group.
F
IGURE
3. Volume of training: the number of sets performed
(per muscle group) during each workout.
Very recently, there have been 2 meta-analyses that
have examined and confirmed the existence of a theoret-
ical continuum for strength-training involvement (19, 21).
These findings, which were determined by using data
from nearly 200 studies, identified a dose-response rela-
tionship for the continuum of training intensities, fre-
quencies, and volumes, and resultant strength increases
among untrained, trained, and athlete populations (Fig-
ures 1–3). Specifically, ES data demonstrated that, for
untrained individuals, maximal strength gains are elic-
ited at a mean training intensity of 60% of 1 repetition
maximum (1RM), 3 days per week, and with a mean
training volume of 4 sets per muscle group. Recreation-
ally trained nonathletes exhibit maximal strength gains
with a mean training intensity of 80% of 1RM, 2 days per
week, and a mean volume of 4 sets. For athlete popula-
tions, maximal strength gains are elicited at a mean
training intensity of 85% of 1RM, 2 days per week, and
with a mean training volume of 8 sets per muscle group.
U
NDERSTANDING THE
D
ATA
:A
PPLICATION OF
THE
D
OSE
-R
ESPONSE
Essentially, this research has confirmed a quantifiable,
functional framework whereby the human musculoskel-
etal system adapts to strength training. Of foremost im-
portance, current data verifies the existence of a dose-
response relationship between training stimulus and
muscular strength adaptation. Accordingly, single-set
and/or very low-volume resistance-training philosophies
may no longer be considered adequate for individuals
seeking improvement in strength beyond that of general
muscular fitness. Significantly greater adaptations are
achievable with subsequent sets of training, even for un-
trained populations (21). Further, and in support of the
recent progression model (14), dose-response data dem-
onstrate that muscular strength development requires
progressive training, such that smaller doses will elicit
greater muscular strength improvement for individuals of
inferior muscular fitness than are needed for more highly
trained individuals. The effort-to-benefit ratio of resis-
tance training, and associated muscular fitness adapta-
tion, is, therefore, highly contingent on training status
(i.e., current strength level and strength-training histo-
ry). Subsequent training to promote muscular strength
improvement must be prescribed on an individualized ba-
sis.
Secondly, the rate of improvement in muscular
strength, after initiation of a given training prescription,
decreases with increased training experience and current
level of muscle conditioning. Faster rates of muscular
strength improvement are typical during earlier periods
of training, especially for previously untrained individu-
als, and are likely attributed to neural adaptations, re-
sulting in enhanced motor unit activation (2, 24). More-
over, there is a diminished return in sheer magnitude of
muscular strength adaptation that accompanies greater
levels of muscular development. A novice resistance
trainer may have the ability to make vast improvements
in short amounts of time, with minimal amounts of train-
ing. Alternatively, a recreational athlete with several
years of training experience will not only make less im-
provement in muscular strength, but the improvements
will be generated more slowly, and the necessary dose of
training to maximize this progress must be subsequently
greater. Even a well-trained recreational athlete, how-
ever, will be capable of greater strength gains, per unit
of time, than would a very experienced, very strong lifter
(Figure 4).
Until now, there has been a paucity in the research
and literature of quantification of the construct of this
phenomenon. The current meta-analytical investigations
were not intended to determine the details of this decre-
ment in training adaptation, but rather, to quantify the
A
PPLICATIONS OF THE
D
OSE
-R
ESPONSE
953
F
IGURE
4. Theoretical dose-response progression continuum.
theorized progression continuum by establishing the nec-
essary mean doses to maximize strength development for
various populations. To use this data, it is important that
the professional understand how to convert these findings
to practical entities. Just as an improper prescription of
medication may have undesirable effects in a medical set-
ting, so too will the improper dosage of training result in
failure of the strength-training program to elicit the de-
sired muscular developmental effect. Clearly, knowledge
of how much training is needed to elicit a desired or need-
ed strength improvement would inevitably serve as a
valuable point of reference for program design. Most prac-
titioners regard the processes of exercise prescription as
a technical, although artistic, endeavor. Accordingly,
skillful professionals may use the dose-response data to
construct the framework for exercise prescription that, ul-
timately, may ensure the continued efficiency and effec-
tiveness of a multi-dimensional program.
Although the meta-analyses have greatly reduced the
subjectivity of exercise prescription for muscular strength
development, the equivocal task of establishing appropri-
ate training doses among specific individuals still exists.
An overprescribed protocol will, at minimum, negatively
affect the training-to-adaptation (effort-to-benefit) ratio.
Unless adjusted, overprescription of the training vari-
ables may eventually facilitate strength-gain plateaus or
even lead to muscular strength decrements, and cause
adverse side effects associated with overreaching and/or
overtraining (11). The design of a program must incor-
porate a systematic method of testing/determining train-
ing status and muscular development to ensure proper
progression and health. In effect, using the dose-response
data may actually demand more time, because careful
scrutiny of training progression is constantly necessary.
On the contrary, under prescription of training may
be appropriately regarded as training inefficiency. Thus,
underprescription becomes less detrimental, or even ir-
relevant, to most nonathlete populations, than it is for
high-level amateur and professional athletes bound by
the time constraints of a sport season or the preparation
phases of training. Program efficiency for sport condition-
ing is, therefore, a measure of developmental effect per
unit of time. Maximizing efficiency is imperative, because
numerous training objectives are set forth during an al-
lotted phase to facilitate safe, successful sport perfor-
mance. The consummate professional must be cautious in
their approach to maximize muscular strength-training
efficiency, because the proper prescription is a highly in-
dividualistic entity.
O
PERATIONAL
D
EFINITIONS
A respective operational assignment is warranted in fur-
ther discussing the dose-response relationship for these
training variables. Most notably, the designations of
training intensity and volume tend to be consistently in-
congruent among professionals. Intensity of training re-
fers to the percentage of 1RM used for a given exercise.
Progressive muscular conditioning necessitates that
greater mean intensities of training are necessary to max-
imize development of strength. For further clarification,
the number of repetitions performed should not be con-
sidered a component of training volume, because inten-
sity should dictate repetitions, not vice versa. This oper-
ational definition for training intensity generates an ob-
jective, quantifiable unit, which is contrary to the more
subjective measure of training fatigue, often exploited in
low-volume programs (8). On a related note, deliberate,
slow contraction speed (to accentuate fatigue) is contra-
indicated for maximal strength development because this
method incorporates absolute training intensities lower
than necessary for progressive strength adaptation (13).
Training frequency may be defined as the occurrence
per unit of time (e.g., calendar week) that a given major
muscle group, or prime mover, is trained. Generally,
higher-volume training dictates the need to partition a
given training program into smaller, less indistinct pro-
grams to accommodate greater time requirements. Full-
body training, which is often prescribed 2 to 3 days per
week, may be reduced to 2 upper-body and 2 lower-body
training sessions per week (4 total sessions). Subsequent-
ly, an upper-body and lower-body split regimen may be
further reduced to target a single muscle or movement in
a given training session, twice a week (5–6 total sessions
per week for all musculature).
Volume of training refers to the number of sets per-
formed per muscle group, per workout. For example, 5
sets of bench press and 4 sets of incline bench press is
equivalent to 9 total sets of training to stimulate strength
adaptation of the respective prime-mover musculature.
Certainly, all exercises are not created equal regarding
limb position, joint articulation, and/or recruitment of
synergistic/stabilization muscles. However, rather than
defining training volume as the total number of sets per
individual exercise, total number of sets per muscle group
is a more appropriate measure of the absolute stress ap-
plied to a given primary-mover(s). Interestingly, it is ap-
parent that with this designation, many supposed low-
volume training philosophies are actually higher-volume
training practices.
The inevitable confounding variable that may prevent
proper use of the dose-response data is feasibility. High-
er-volume workouts that are more strenuous take more
time, both as individual training sessions and collectively
over time. Many efficiency strategies have been devised
that enable more sets of exercise to be performed in a
training routine, such as circuit training, complex train-
ing, super-setting, drop-setting, tri-setting, etc. Profes-
sionals are expected to understand these strategies and
be able to implement them at the appropriate place and
time. However, there is an unavoidable downfall to most
of these strategies: a trade-off must take place with in-
tensity, accompanying higher training volumes and ulti-
mately, fatigue. As the dose-response data demonstrate,
the basic strength phase calls for a simultaneous increase
954 P
ETERSON
,R
HEA
,
AND
A
LVAR
F
IGURE
5. (a) Absolute strength function curve. (b) Marginal
strength function curve.
in intensity and volume of training. To accomplish both,
rest intervals between sets must be accentuated, and sin-
gle joint, unilateral exercises must become less empha-
sized.
Furthermore, if the goal or necessity of training is to
maximize muscular strength development, the dose-re-
sponse data must be used in conjunction with a plan that
also specifically accommodates muscular endurance, hy-
pertrophy, and muscular power, as well as addresses the
vital skill-related components of fitness, including accel-
eration, speed, balance, coordination, reaction time, and
nonlinear movement. Whether it is for recreational- or
elite-level training, the deliberate fluctuation of training
dosages to pursue specific muscular fitness objectives al-
lows for continued progression over time. In fact, partic-
ipating in a given training regimen that is devoid of var-
iation may well lead to suppression of physiological and
neurophysiological adaptations. It is very important that
the dose-response data refers to mean intensities, fre-
quencies, and volumes. Within a training microcycle, the
daily, or even weekly, values of these doses may be al-
tered, as long as the average remains at the necessary
predetermined level to optimize training effect over the
duration of the immediate strength program/phase.
As with traditional periodization, this manipulation
process typically necessitates that the intensity of train-
ing be gradually increased, and volume decreased, over
time, allowing for adequate opportunity to adapt to the
higher intense training protocol. To concurrently increase
both volume and intensity of training, without overtrain-
ing, it must be done indirectly. Conceptually, volume per-
formed per unit of time may be increased while decreas-
ing the volume performed per training session, simply by
altering training frequency. In other words, there may
still be a trade-off with volume and intensity in a given
training session, as long as there is a subsequent trade-
off between frequency of training per week and volume
per workout.
T
HE
L
AW OF
D
IMINISHING
R
ETURNS
The law of diminishing returns states that as the quan-
tities of an input increase, the resulting rate of output-
increase eventually decreases (12). Often known as the
law of diminishing marginal returns in economics, this
law is used to determine the optimal working relationship
between total and marginal values of a production sys-
tem. Concerning muscular development, we may consider
strength improvement as the product, and dose of train-
ing, the input. The change in strength adaptation asso-
ciated with a 1-unit change in training dosage per unit of
time is, therefore, known as the marginal product of
strength.
Figure 5 represents a hypothetical snapshot of the
continuous relationship between absolute strength, mar-
ginal strength, and training dose for a given individual.
The strength increase per unit of training dose is greatest
at, and denoted by, the inflection point A (the point at
which the slope of the line is greatest) on the marginal
strength curve. It is at this point on the curve that ‘‘In-
creasing Returns’’ peaks and transitions into ‘‘Diminished
Returns.’’ The marginal strength curve in Figure 5 is syn-
onymous with the dose-response trends in that it submits
a quantified magnitude of effect for a specific dose of
training relative to other doses. Muscular strength de-
velopment is maximized at the point A. Point A is also
the point wherein the most efficient training takes place,
because strength change and training dosage are func-
tions that exists per unit of time. If training dosage is
expanded past point A, absolute strength gain continues
to progress, but with a diminished return of effort to ben-
efit. Eventually, the increased diminished return becomes
‘‘Negative Returns,’’ because absolute strength plateaus
at point B and begins to decrease (intersects the zero-
value of marginal strength).
Regarding diminishing returns, a prevalent argument
against progressive training contests the need for addi-
tional volume beyond that of the initial set, citing a
marked drop-off in strength adaptation outcome with any
subsequent training. If, in fact, it is true that the first set
elicits a purported 90% of the potential gains in strength,
multiple sets may seem inherently unnecessary, and ul-
timately not worth the time and effort. Interestingly, with
simple extrapolation of the coded volume data within the
meta-analyses, it is actually possible to examine and
quantify this theory with an objective measuring tool and
to report the data in relation to an overall percentage of
strength gains.
Specifically, as may be seen in Figure 6, when the
data for an athletic population is examined it seems that,
approximately, a mere 26% of strength-gain potential is
elicited with 1 set of training. In fact, dose-response data
suggests that an athlete does not satisfy the requirement
for 75% of strength developmental propensity until an av-
erage of 4 sets of training are completed. Finally, the data
also reveals that for the athlete population, a mean of 8
sets of training per muscle group is a requisite to ensure
maximal strength-training effect. As an important point
of clarification, these data do not confer that a single spe-
cific volume of training be an appropriate recommenda-
tion for every athlete or nonathlete individual, nor should
it encourage the prescription of a particular volume for
an extended time. However, the findings do soundly up-
A
PPLICATIONS OF THE
D
OSE
-R
ESPONSE
955
F
IGURE
6. Using effect size to extrapolate to percent gains in
muscular strength.
F
IGURE
7. Training-to-failure vs. not training-to-failure: im-
plications to volume of training.
hold the conception that a dose-response for strength ad-
aptation exists, and that trends demonstrate a required
progression of training volume to approach an eventual
average of 8 sets of training to accompany progressive
muscular strength adaptation. As a notable corollary,
Figure 6 also demonstrates that virtually no differences
exist between trained and untrained populations as they
respond to variation in volumes of strength training.
Once again, the first set of training is only responsible for
producing approximately 50% of the overall net potential
gain in strength for these 2 populations, needing approx-
imately 4 sets to ultimately achieve the desired maximal
effect. Clearly, both athletes and nonathletes require
multiple sets of strength training to promote optimization
of muscular strength development per unit of time. Low-
volume and single-set protocols drastically restrict the po-
tential strength adaptation of athletes and nonathletes,
respectively.
D
EFENSE OF THE
M
ETA
-A
NALYSIS AND
D
OSE
-
R
ESPONSE
Of course, these meta-analyses (19, 21) and dose-response
trends have not been devoid of critical scrutiny. In fact,
since the publication of the first of these meta-analyses
(21) there have been several consistent areas of concern
regarding the efficacy of the dose-response for muscular
strength program prescription. Accordingly, the following
are several standard arguments that have been used
against the meta-analyses and proposed dose-response
data, as well as the authors’ collective defense.
Adherence
Two frequently cited issues relative to the implementa-
tion of the dose-response data for athlete populations are
increased time commitment per training session, and sub-
sequent training adherence. It has been argued that pro-
grams that approach the upper end of the dose-response
continuum (which, again, suggest 8 sets of training per
muscle group) will surely last longer than 1 hour, and,
thus, inevitably fail to produce desired results. Seeming-
ly, this argument stems from a limited body of research
advising that 60-minute training sessions be the defini-
tive upper limit in time commitment for exercise pro-
grams.
The 1998 ACSM Position Stand states that ‘‘Programs
lasting more than 60 minutes per session appear to be
associated with higher dropout rates’’ (Ref. 1, p. 983). A
critical question must be raised: will strength adaptation
and adherence diminish if training sessions last longer
than this seemingly simplistic 60-minute rule? The an-
swer to this question becomes less definitive if we consid-
er athletic populations of higher training status. Athletes
are truly different than the general populations regarding
adherence concerns. However, when taking the athletic
population out of the equation, it is interesting to note
that the 1998 Position Stand references a study that
shows that the average time required to complete a 3-set
resistance-training program is approximately 50 minutes
(1, 18). Thus, for trained and untrained nonathletes, it is
highly probable that a training session may be accom-
plished in 1 hour, especially if the recommended training
volume for trained and untrained populations is used at
an average of 4 sets per muscle group.
In reality, there is an upper limit to the amount of
time athletes can dedicate to strength-training programs.
The true limit is an individualistic entity, however, and
it becomes the tedious job of the strength and condition-
ing professional to use time to optimize the conditioning
benefits for each athlete. Motivational factors must be
considered for this population. One global and resounding
motivational factor is evident as positive results of the
actual strength-training intervention. If we understand
the theory that governs the dose-response, and what it
takes to optimize muscular strength adaptation, credence
is lent to additional intensity and volume in program pre-
scription. One of the prime motivators in sport perfor-
mance training is to produce the desired result of maxi-
mal strength improvement. The dose-response offers sci-
entific credibility to the design and implementation of ap-
propriate frequency, volume, and intensity of program
prescription.
An additional means of increasing strength-training
adherence would be through the processes of systematic
goal setting. Several renowned psychologists and sport
psychologists have discovered that specific and challeng-
ing goals lead to higher intrinsic motivation and perfor-
mance outcomes than easily attainable goals (15–17, 22).
Further, this investigative rationale has also recently
been advocated for use in sport performance programs
(26). If we are to follow the research and postulated ap-
plications, it is evident that multiple-set programs may
be used as given explicit, challenging process-oriented
goals. Program design may rely on systematic increases
956 P
ETERSON
,R
HEA
,
AND
A
LVAR
in volume and intensity of training as additional moti-
vational factors. To do this, specific alterations in the in-
tensity and volume may be computed and, thus, the train-
ing sessions (i.e., process) become part of the goals and
ultimate strength development (i.e., product).
Training-to-Failure
One of the most vehement arguments against the dose-
response for strength training has come from high-inten-
sity advocates. Accordingly, as theory would suggest, a
high-intensity (i.e., single set or very low volume often
combined with deliberate, low movement velocity) pro-
gram can be just as effective as a multiple-set program if
sets are performed to the point of absolute muscular fail-
ure (8). However, again, with simple extrapolation of
training-to-failure data from the meta-analyses, a small
sample of data and ES were analyzed (n575 studies).
Each study had to include training-to-failure as a com-
ponent of the overall data set, and control for the mag-
nitude of training frequency and intensity.
As can be seen in the Figure 7, training-to-failure does
not elicit greater gains than not training-to-failure. This
is true even when multiple sets of training-to-failure are
compared with training sessions that are not training-to-
failure.
The Meta-Analyses
Lastly, it is necessary to briefly address the content of
several recently published reports on the Internet (23, 27)
that have appraised the value of the meta-analyses. Cer-
tainly, these critiques were reasonably predictable be-
cause there is a long-standing discord between groups of
conflicting strength-training philosophies. Unfortunately,
whereas dissimilar opinions and viewpoints should, and
often can, foster development and refinement of the sci-
ences and practices within a field, from time to time, in-
novative research inquiry is touted as undisciplined and/
or ‘‘questionable’’ if findings are not in support of a par-
ticular individual’s or group’s agenda. Specifically, there
are several notorious, self-appointed experts who have
made a name by aimlessly denigrating the work of others
while supporting an obstinate bias, with critiques that
lack scrutiny, justification, and accuracy.
The majority of the initial critique (27) focuses on
methodological issues relating to the meta-analysis.
Methodological debates have existed since the creation of
meta-analytic procedures, as have existed regarding
methodologies of narrative reviews and experimental re-
search. Any suggestion that there is a single correct
method for conducting a meta-analysis is as imprudent
as suggesting that there is only 1 correct method for con-
ducting experimental research. In fact, the checklist
wherein a given meta-analysis is evaluated is highly var-
iable and is dependant on whichever of the nearly 100
resources that scrutinize the topic of meta-analytic pro-
cedure for the various social, behavioral, and health sci-
ences is chosen. Any suggestion that the findings of the
analyses are of ‘‘little value’’ as a result of not including
some of the methodological steps that this particular re-
viewer presents and references is extremely shortsighted.
Although we have, as conscientious investigators, ac-
knowledged that each of our analyses has its respective
limitations, the critique in question provides no quanti-
tative evidence that the methodological criticisms would
actually have any influence on the findings and conclu-
sions presented.
The following are several examples of uncalculated ob-
servations. Ultimately, dissection of this critique from Dr.
Winnett (27) actually helps to demonstrate the soundness
with which the meta-analyses substantiate a dose-re-
sponse for strength training. The reviewers reported that
the 2 meta-analyses proclaimed to have analyzed ‘‘all’’ ex-
isting strength-training research articles, to date (Ref. 27,
p. 12; Ref. 23, p. 56). First, there have been no such claims
made wherein ‘‘all’’ studies examining strength develop-
ment were secured; the only claim was to have conducted
a search for studies including a strength-training inter-
vention. Indeed, many more studies have been conducted
on strength development than were included in the meta-
analyses. Securing every study ever conducted would be
unreasonable, because countless more exist. However, we
have continued to add studies to the database as we se-
cure them, and have yet to observe any alteration in the
dose-response trends presented in our published analyses
(19, 21). Based on the size of our current database (.250
studies), it would take literally thousands of ES, all show-
ing drastically different strength outcomes than the stud-
ies included in the meta-analyses to have a significant
impact on the proposed dose-response data. Such a situ-
ation is unlikely to exist.
The analyses were also criticized for the inclusion of
multiple ES from a single study, suggesting again that
such methodological issues should result in rejection of
the findings presented. This is an issue that has received
significant discussion and debate in meta-analytic
groups, and, for the most part, there has been an accep-
tance of the inclusion of multiple ES from each individual
study, especially when large numbers of studies are in-
cluded in the analysis. If, for instance, a small number of
studies were included, and 1 or 2 of those studies included
many ES, then those studies would have a greater impact
on the findings of the analysis. However, our analyses
included 37 (19), and 140 (21) studies, respectively, with
very few studies only contributing a single ES. Most stud-
ies reviewed included multiple, dissimilar measures of
strength (i.e., squat, bench press, etc.). There is no reason
to expect, based on the large number of studies included,
that our findings were in any way biased by such meth-
ods. Further, once again the reviewer has provided no
evidence to otherwise support the claim that such meth-
odological issues render the findings worthless.
Several additional criticisms were raised regarding
our statistical analyses. Shockingly, there were consid-
erable concerns regarding the findings for training vol-
ume. By performing a simple t-test of 2 sets vs. 4 sets,
the reviewer concluded that 4 sets do not elicit greater
strength gains because a nonsignificant pvalue was ob-
tained. However, sole reliance on probability values to
make statistical decisions has received a great deal of
criticism among researchers recently, because numerous
potential errors arise, especially in low-power studies
(25). One of the benefits of calculating, and combining ES,
is the ability to examine the actual magnitude of the
treatment effect instead of relying solely on pvalues to
examine differences between treatments. The use of the
reviewer’s t-test (and nonsignificant pvalue) was the only
source of evidence to support any of the claims made that
prompted the conclusion that the meta-analyses do not
support the performance of multiple-set training. Inter-
A
PPLICATIONS OF THE
D
OSE
-R
ESPONSE
957
estingly, this t-test was performed between 2 sets and 4
sets—both of which represent multiple-set programs.
When analysis of covariance is performed, with intensity
and frequency as covariates on the different volumes of
training, it is strongly demonstrated that multiple sets of
training are more effective at eliciting strength improve-
ments than single-set training. Further, whereas 2-, 3-,
and 4-set programs did not result in significantly differ-
ent effects sizes (p.0.05), it should be noted that, in a
number of the studies within the analyses, control groups
(no training at all) demonstrated improvements in
strength (ES 50.30), which, based solely on probability
values, were found to be similar to the amount of strength
improvement with 1 set of training (p.0.05). Therefore,
if reliance on probability values is the only source of valid
information achieved by the research process (as sug-
gested by opponents of the meta-analyses we have con-
ducted), then those same pvalues reject the notion that
single-set resistance training is any better than no train-
ing at all and do not result in the same degree of strength
development as multiple-set programs.
Finally, numerous blatant discrepancies between ac-
tual facts from the published analyses and allegations
found in the critiques must be exposed. First, the meta-
analysis on muscular strength for competitive athletes
(19) is not a ‘‘subset’’ of the previously published meta-
analysis on untrained and trained populations (21), as
was suggested (Ref. 27, p. 16). Between the 2 analyses,
less than 10 references contributed to both analyses,
whereas nearly 200 references in total—with a combined
1,800 ES—were used. A simple reference check could
have pinpointed this fact and alleviated any related con-
fusion. Furthermore, inclusion criteria for the 2 meta-
analyses dictated a rigorous elimination process, whereby
more than 1,000 articles were searched for and coded by
the primary authors, during the course of almost 4 years.
Each research article was expected to contain a controlled
strength-training intervention and include a detailed de-
scription of the following information for viable inclusion
in the meta-analyses: (a) number of participants (n); (b)
duration of training (i.e., number of weeks for the train-
ing intervention); (c) periodization (coded as: 1 5tradi-
tional linear periodization, 2 5undulating periodization,
35no variation-progressive resistance, and 4 5nonper-
iodized); (d) mode (coded as 1 5free weights, 2 5ma-
chines, and 3 5both); (e) volume of training (coded as
sets per muscle group per workout); (f) intensity of train-
ing (coded as percent of 1RM); (g) frequency of training
(coded as number of days trained per week per muscle
group); (h) training-to-failure (coded as 1 5yes, 2 5no,
and 3 5not specified); (i) status (coded as 1 5untrained:
less than 1 year of consistent strength training, 2 5re-
creationally trained: more than 1 year of consistent
strength training or high school athlete, and 3 5college,
professional, and/or elite athlete); (j) creatine use (coded
as 1 5yes, and 2 5no); (k) sex (coded as 1 5male, and
25female); (l) age (coded for mean age); (m) preinter-
vention and postintervention strength measures (coded
as means and standard deviation descriptive statistics).
Additionally, in contrast to the claim that we failed to
address the issue of differing numbers of ES among vary-
ing training volumes and intensities, each of the papers
actually contained a clear cautionary statement. In fact,
we have pointed out several times throughout our anal-
yses that this limitation should be noted, and that, if all
levels were equated, the dose-response trends may have
been different. However, as previously mentioned, we
have been consistently adding articles and ES calcula-
tions to our databases, with no tangible alterations to
date. Should there be a change, it would certainly be our
duty—as ethical investigators—to present that informa-
tion. Further, and related to this issue, the review claims
that we have neglected to present any physiological or
theoretical rationale to explain the data and conclusions.
Paradoxically, in every published paper, we have dis-
cussed the concepts of overload and progression, two
physiological principles of training adaptation that sup-
port, and are supported by, our analyses, whereas oppo-
nents of these principles (i.e., single-set advocates) con-
tinually fail to present a physiological rationale for the
opinion that single-set training will elicit maximal
strength gains.
Without doubt, the field of exercise and sport sciences
will continue to evolve only if there is constant critical
examination of the knowledge disciplines and methods
that govern the respective research. Nevertheless, if an
impact is to be made on the rising number of profession-
als that administer exercise and sport-conditioning pre-
scriptions, and more importantly, the general consumer
who seeks education in pursuit of fitness aspirations,
there must be a shift in paradigm to acquiesce to ‘‘com-
mon goals’’ by means of ‘‘different roles.’’ Much of the clin-
ical credibility lost during the early part of the last 40
years, with the emergence of the first scientific-reminis-
cent literature on strength training, has been regained.
Relatively speaking, however, the science behind the
principles of strength training and disciplines of sport
conditioning is still very much in an infancy stage. As a
result, the professional frontlines of the field have, by ne-
cessity, preceded the scientific community for quite some
time. Hence, any all-encompassing blanket arguments
against the principles that are advocated by the National
Strength and Conditioning Association and the American
College of Sports Medicine, because of a ‘‘lack of empirical
support’’ (23), should be viewed merely as spiteful. More-
over, with the availability of today’s training technologies
and methodologies, a persistent reliance on age-old be-
liefs that ignore valuable and innovative empirical evi-
dences, so as to oversimplify and over generalize
strength-training prescription, is clearly illogical.
P
RACTICAL
A
PPLICATIONS
A principal objective in sport conditioning includes max-
imizing training effect per unit of time to allow multiple
goals to be met simultaneously. Current dose-response
data establishes the point at which marginal strength is
maximized for various inputs (i.e., volume, intensity, and
frequency). It should be a primary objective of all profes-
sionals to determine the point at which a given athlete,
or group of athletes, maximizes progress. The collabora-
tive results of the 2 meta-analyses demonstrate that de-
velopment of muscular strength through resistive exer-
cise may be optimized through careful prescription and
manipulation of the training variables, such that distinct
dose-response relationships correspond to specific popu-
lations. Furthermore, training status (i.e., training his-
tory and level of current muscular fitness) directly dic-
tates these optimal doses and influences the potential
magnitudes of the responses. In particular, for healthy
individuals seeking to experience muscular strength de-
958 P
ETERSON
,R
HEA
,
AND
A
LVAR
velopment beyond that of general health, a requisite in-
crease in resistance-training dosage must accompany in-
creases in training time and experience. These investi-
gations further demonstrate that the effort-to-benefit ra-
tio is different for untrained, recreationally trained, and
athlete populations; thus, emphasizing the necessity of
appropriate exercise prescription to optimize training ad-
aptation. The meta-analytic procedure is a scientific, rig-
orous, objective, and quantifiable synthesis of a given
treatment and resultant treatment effect. Relying on such
scientific evidence, combined with professional experience
and competency, may result in the most effective and re-
liable training prescriptions.
R
EFERENCES
1. A
MERICAN
C
OLLEGE OF
S
PORTS
M
EDICINE
. The recommended
quantity and quality of exercise for developing and maintaining
cardiorespiratory and muscular fitness, and flexibility in
healthy adults: Position stand. Med. Sci. Sports Exerc. 30:975–
991. 1998.
2. B
EHM
, D.G. Neuromuscular implications and applications of
resistance training. J. Strength Cond. Res. 9:264–274. 1995.
3. B
YRD
, R. Strength training: Single versus multiple sets. Sports
Med. 27:409–416. 1999.
4. C
ARPINELLI
, R.N.,
AND
R.M. O
TTO
. Strength training—Single
versus multiple sets. Sports Med. 26:73–84. 1998.
5. C
ARPINELLI
, R.N.,
AND
R.M. O
TTO
. Strength training: Single
versus multiple sets—The authors’ reply. Sports Med. 27:412–
416. 1999.
6. C
ARPINELLI
, R.N., R.M. O
TTO
,
AND
R.A. W
INETT
. A critical
analysis of the ACSM position stand on resistance training:
insufficient evidence to support recommended training proto-
cols. J. E. P. Online. 7:1–60. 2004.
7. C
OHEN
, J. Things I have learned (so far). Am. Psych. 45:1309.
1990.
8. D
ARDEN
, E. The new high intensity training. New York, NY:
Rodale, 2004.
9. D
E
H
OYOS
,D.
AND
M.L. P
OLLOCK
. Single set vs. multiple sets
for adult fitness. I. D. E. A. Personal Trainer. May:7–9. 1998.
10. F
EIGENBAUM
, M.S.,
AND
M.L. P
OLLOCK
. Prescription of resis-
tance training for health and disease. Med. Sci. Sports Exerc.
31:38–45. 1999.
11. H
ALSON
, S.L.,
AND
A.E. J
EUKENDRUP
. Does overtraining exist?:
An analysis of overreaching and overtraining research. Sports
Med. 34:967–981. 2004.
12. H
IRSCHEY
,M.Fundamentals of Managerial Economics. (7th
ed.). Mason, OH: Thomson South-Western. 2003.
13. K
EELER
, L.K, L.H. F
INKELSTEIN
,W.M
ILLER
,
AND
B. F
ERNHALL
.
Early-phase adaptations of traditional-speed vs. superslow re-
sistance training on strength and aerobic capacity in sedentary
individuals. J. Strength Cond. Res. 15:309–314. 2001.
14. K
RAEMER
, W.J., K. A
DAMS
,E.C
AFARELLI
, G.A. D
UDLEY
,C.
D
OOLY
, M.S. F
EIGENBAUM
, S.J. F
LECK
,B.F
RANKLIN
, A.C. F
RY
,
J.R. H
OFFMAN
, R.U. N
EWTON
,J.P
OTTEIGER
, M.H. S
TONE
, N.A.
R
ATAMESS
,
AND
T. T
RIPLETT
-M
CBRIDE
. American College of
Sports Medicine position stand. Progression models in resis-
tance training for healthy adults. Med. Sci. Sports Exerc. 34:
364–380. 2002.
15. K
YLLO
, L.G.L.,
AND
D.M. L
ANDERS
. Goal setting in sport and
exercise: A research synthesis to resolve the controversy. J.
Sport Exerc. Psychol. 17:117–137. 1995.
16. L
OCKE
, E.A.,
AND
G.P. L
ATHAM
. Building a practically useful
theory of goal setting and task motivation. A 35-year odyssey.
American Psych. 57:705–717. 2002.
17. L
OCKE
, E.A., K.N. S
HAW
, L.M. S
HAARI
,
AND
G.P. L
ATHAM
. Goal
setting and task performance: 1969–1980. Psychol. Bull. 90:
125–152. 1981.
18. M
ESSIER
, J.P.,
AND
M. D
ILL
. Alteration in strength and maxi-
mal oxygen uptake consequent to nautilus circuit weight train-
ing. Res. Q. Exerc. Sport 56:345–351. 1985.
19. P
ETERSON
, M.D., M.R. R
HEA
,
AND
B.A. A
LVAR
. Maximizing
strength development in athletes: A meta-analysis to deter-
mine the dose-response relationship. J. Strength Cond. Res. 18:
377–382. 2004.
20. R
HEA
, M.R. Synthesizing strength and conditioning research:
The meta-analysis. J. Strength Cond. Res. 18:921–923. 2004.
21. R
HEA
, M.R., B.A. A
LVAR
, L.N. B
URKETT
,
AND
S.B. B
ALL
.A
meta-analysis to determine the dose-response for strength de-
velopment. Med. Sci. Sports Exerc. 35:456–464. 2003.
22. R
YAN
, R.M.
AND
E.L. D
ECI
. Self-determination theory and the
facilitation of intrinsic motivation, social development, and
well-being. Am. Psychol. 55:68–78. 2000.
23. S
MITH
, D.,
AND
S. B
RUCE
-L
OW
. Strength training methods and
the work of Arthur Jones. J. E. P. Online. 7:52–68.
24. S
TARON
, R.S., D.L. K
ARAPONDO
, W.J. K
RAEMER
, A.C. F
RY
, S.E.
G
ORDON
, J.E. F
ALKEL
, F.C. H
AGERMAN
,
AND
R. S. H
IKIDA
. Skel-
etal muscle adaptations during early phase of heavy-resistance
training in men and women. J. Appl. Physiol. 76:1247–1255.
1994.
25. T
HOMAS
, J.R., M.R. L
OCHBAUM
, D.M. L
ANDERS
,
AND
C. H
E
.
Planning significant and meaningful research in exercise sci-
ence: Estimating sample size. Res. Q. Exerc. Sport. 68:33–43.
1997.
26. T
OD
, D.,
AND
M. M
CGUIGAN
. Maximizing strength training
through goal setting. J. Strength Cond. 23:22–27. 2001.
27. W
INNETT
, R.A. Meta-analyses do not support performance of
multiple sets or high volume resistance training. J. E. P. On-
line. 7:10–20. 2004.
Address correspondence to Mark D. Peterson, mdpeterz@
hotmail.com.
