Sunteți pe pagina 1din 22

Sports Med 2005; 35 (3): 213-234

REVIEW ARTICLE 0112-1642/05/0003-0213/$34.95/0

 2005 Adis Data Information BV. All rights reserved.

Challenges in Understanding the


Influence of Maximal Power Training
on Improving Athletic Performance
John Cronin1 and Gord Sleivert2
1 New Zealand Institute of Sport and Recreation Research, Auckland University of Technology,
Auckland, New Zealand
2 Human Performance Laboratory, Faculty of Kinesiology, University of New Brunswick,
Fredericton, New Brunswick, Canada

Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
1. Seminal Training Practice and Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
2. Power-Load Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
2.1 Upper Body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
2.2 Lower Body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
3. Power and Performance: Cross-Sectional Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
3.1 Upper Body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.2 Lower Body . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.2.1 Cyclic versus Acyclic Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.2.2 Absolute versus Relative Power Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
3.2.3 Maximum Power versus Power Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
3.2.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
4. Power and Performance: Training Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232

Abstract The ability to optimise muscular power output is considered fundamental to


successful performance of many athletic and sporting activities. Consequently, a
great deal of research has investigated methods to improve power output and its
transference to athletic performance. One issue that makes comparisons between
studies difficult is the different modes of dynamometry (isometric, isokinetic and
isoinertial) used to measure strength and power. However, it is recognised that
isokinetic and isometric assessment bear little resemblance to the accelerative/
decelerative motion implicit in limb movement during resistance training and
sporting performance. Furthermore, most people who train to increase power
would have limited or no access to isometric and/or isokinetic dynamometry. It is
for these reasons and for the sake of brevity that the findings of isoinertial
(constant gravitational load) research will provide the focus of much of the
discussion in this review.
One variable that is considered important in increasing power and performance
in explosive tasks such as running and jumping is the training load that maximises
214 Cronin & Sleivert

the mechanical power output (Pmax) of muscle. However, there are discrepancies
in the research as to which load maximises power output during various resistance
exercises and whether training at Pmax improves functional performance is
debatable. There is also some evidence suggesting that Pmax is affected by the
training status of the individuals; however, other strength variables could quite
possibly be of greater importance for improving functional performance. If Pmax
is found to be important in improving athletic performance, then each individual’s
Pmax needs to be determined and they then train at this load. The predilection of
research to train all subjects at one load (e.g. 30% one repetition maximum
[1RM]) is fundamentally flawed due to inter-individual Pmax differences, which
may be ascribed to factors such as training status (strength level) and the exercise
(muscle groups) used. Pmax needs to be constantly monitored and adjusted as
research suggests that it is transient. In terms of training studies, experienced
subjects should be used, volume equated and the outcome measures clearly
defined and measured (i.e. mean power and/or peak power). Sport scientists are
urged to formulate research designs that result in meaningful and practical
information that assists coaches and strength and conditioning practitioners in the
development of their athletes.

1. Seminal Training Practice 1RM. Higher tensions would inhibit the ability of
and Research muscles to move quickly, which was thought a
fundamental prerequisite of power training.
Power can be defined as the amount of work
produced per unit time or the product of force and In contrast, some coaches have thought heavier
velocity. The development of power and its transfer- loads are necessary for improved power production.
ence to performance has been the source of interest Poprawski[3] compared the strength power results of
and discussion for years. Initially, coaches and Edward Sarul, 1983 World Champion in the shot-
strength and conditioning practitioners debated the put, to nine well trained shot-putters, which included
merits of using various loads for the development of >21m throwers. While Sarul was slightly stronger
power. From the literature there appeared two than the group average in the bench press (1.4%),
schools of thought, one that was Western in origin snatch (7.9%), power clean (5.3%) and squat
and espoused the use of lighter loads (<50% one (7.8%), the major differences occurred in tests of
repetition maximum [1RM]) for improving power speed and power at heavy loads in those respective
output and athletic performance, whereas Eastern exercises. For example, Sarul’s snatch velocity
bloc coaches and trainers proposed that heavier ranged from 4.13% faster at 20kg than the average
loads (50–70% 1RM) were superior. For example, to 22.13% faster at 80kg. Similarly, his squat veloc-
Counsilman[1] a sport scientist and swimming coach, ity was 2.74% greater at 40kg but 25.71% greater at
argued that athletes needed to move light loads at 140kg. These findings led Poprawski[3] to conclude
high speed, as fast movements activated the fast that movement velocities at higher loads (50–70%
fibres. Conversely, slow training recruited fibres 1RM) were critical determinants of athletic success
with slow contraction characteristics, which was in athletes and training emphasis should be placed
thought counter-productive to power training. Simi- on moving lighter loads (50% 1RM) quickly, rather
larly, Behm,[2] in discussing the use of surgical than just striving to lift more weight. Spassov[4] in a
tubing for a tennis power programme, suggested that discussion of programme design for athletes, stated
this should be combined with traditional weight that experts believe loads of 50–70% 1RM per-
training incorporating loads of not more than 50% formed at a maximal rate, develop explosive power.

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
Maximal Power Training and Improving Athletic Performance 215

Verkhoshansky and Lazarev[5] in a discussion of example, Edgerton et al.,[11] discussing the effects of
Eastern bloc principles for the training of speed and muscle architecture, depicted maximum power out-
strength believed also that loads of 50–70% 1RM put (Pmax) occurring at a range of isometric forces
were necessary for the development of ‘explosive depending whether the knee extensors (45%), knee
strength’. Tidow[6] suggested that heavy loads might flexors (59%), plantar flexors (35%) and dorsiflex-
be equally as effective as light loads for stimulating ors (53%) were studied. These differences were
fast motor unit activity, as the fastest high threshold attributed to different fibre lengths that make deter-
units need to be recruited to lift heavy loads. This mination of Pmax difficult, as Pmax is related to fibre
dichotomy as to which loads (light vs heavy) best shortening velocity, which in turn is related to the
maximise power development remains topical and is number of sarcomeres arranged in series (fibre
still the source of much research, but clearly there length). In fact, Edgerton et al.[11] concluded that the
was a dichotomy of opinion between Eastern and complication of variable fibre lengths makes any
Western bloc power training philosophies. conclusion regarding power per unit of muscle
weight per muscle group of limited value. It would
One approach to solving this dichotomy is to
seem some of the previous assumptions made by
study the relationship between force and velocity,
other authors, based on these research findings, are
since power is the product of both these variables. It
misplaced. Additionally, it would seem that study-
is well known that as load increases, the force output
ing the power output of whole muscle in vivo would
of muscle in concentric contractions increases with a
have greater practical significance to athletes,
concomitant decrease in the velocity of shortening.
coaches and trainers. The power outputs across a
This phenomenon is known as the force-velocity
spectrum of loads (power-load spectrum) using dy-
relationship of muscle.[7] It is thought that maximum
namic multiarticular exercises similar to those used
power output is the product of optimum force and
during weight training need to be examined, the
optimum shortening velocity. For isoinertial con-
results of which should give a greater appreciation
tractions, it has been suggested that maximum pow-
of the load that maximises mechanical power output
er output occurs at approximately 30% of maximum
in a functional context.
shortening velocity or at approximately 30% of
maximum isometric force.[7,8]
2. Power-Load Spectrum
Many researchers have endorsed such loading for
maximising power output[9,10] citing the research of When studying the power-load relationship, one
Edgerton et al.,[11] Faulkner et al.[12] and Kaneko et must be cautious of extrapolating findings from the
al.[13] as support for the utilisation of such loads. literature since some research has investigated the
However, closer scrutiny of this research leaves one power-load relationship indirectly. That is, the rela-
thinking that such conclusions are somewhat mis- tionship between load and power has been investi-
leading. For example, Faulkner et al.[12] certainly gated, but the load that maximised power output was
reported peak powers at approximately one-third of not reported. For example, using subjects from a
maximal shortening velocity; however, they do not weight-training class, Mastropaolo[14] measured
state at what relative force peak power output occurs power output across loads of 20–100% 1RM. It was
(the reader having to extrapolate this information reported that subjects were tested using a bench-
from the graphs provided). Whether power differs press motion, although the figure depicting the exer-
significantly across the power-force spectrum is, cise appears to be a shoulder-press machine. None-
therefore, unclear. In fact, the power profile of slow theless, power profiles based on this movement are
and mixed muscles appears similar across loads of detailed in graphical form and the authors concluded
15–50% maximum isometric force.[12] Furthermore, (without any apparent statistical support) that the
whether such findings are applicable to whole mus- load maximising power output occurred at 40%
cle or biarticular movement is questionable. For 1RM. However, loads from 40% to 60% 1RM ap-

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
216 Cronin & Sleivert

pear very similar in power output. Making such in table I. In all cases, derivatives of the bench press
interpretations without statistical analysis is prob- have been used to study the power-load relationship.
lematic. As can be observed from table I, the loads reported
The findings of another often-cited paper that to optimise Pmax are similar irrespective of subject
investigated the power-load relationship[13] also training status, age and the type of bench-press
need to be interpreted with caution. In this study, 20 motion used. Most studies report a band in which
male subjects were allotted to four training groups Pmax occurs, some research also indicating that
based on their maximum isometric strength. They loads either side of this band are not significantly
trained their elbow flexors using either isotonic (0%, different.[15,16] It would seem then that the majority
30% or 60%) or isometric (100%) contractions. of research reports loads of 30–70% 1RM as the
Peak power of the elbow flexors during concentric intensities that maximise mean and peak power out-
muscle actions was observed at intermediate move- put.
ment velocities of approximately 30% of maximum Three observations from table I appear notewor-
shortening velocity and 30% of maximum isometric thy. First, greater power outputs are associated with
strength.[13] These authors justifiably chose to ex-
the professional and semi-professional rugby league
amine the effects of three loads on Pmax. However,
players, which is no doubt a function of their greater
such a design does not mean that 30% 1RM is the
body mass, training status (maximal strength) and,
load that maximises power output. The load that
therefore, greater relative loads used for the calcula-
maximised power output could be anywhere be-
tion of Pmax. For example, Baker et al.[15] reported
tween 30–60% of maximum isometric force, yet
mean body mass and 1RM of 92.0 ± 11.1kg and
many authors[9,10] continue to cite this study as sup-
129.7 ± 14.3kg, respectively, whereas Cronin et
port for light loads (30%) producing maximal
al.[16] reported 89 ± 2.5kg and 86.3 ± 13.7kg, respec-
mechanical power output. Furthermore, uniarticular
tively, for club rugby players. Nonetheless the band
motion was examined in this study and untrained
of loads that maximised mean power output was
subjects were used, which limits generalisability to
very similar, although ironically it appears that the
athletic populations. Also, maximal mechanical dy-
Pmax of the better trained (greater maximal strength)
namic power output was reported based on a per-
rugby league players occurred at a lower percentage
centage of maximal isometric force with no esti-
of their 1RM. Secondly, it may be that the load that
mates of power in relation to the actual dynamic
maximises peak power is slightly greater than the
exercises used in strength training (e.g. squat and
load that maximises mean power output. The find-
bench press) or the athletic performance itself. It is
ings for the lower body (see table II) would certainly
quite likely that the force at which Pmax occurs
support this contention. Thirdly, and related to the
differs, if expressed relative to a dynamic strength
first point, it would appear that Pmax may be tran-
measure (% 1RM). Based on these observations, it
sient and is affected by the strength status of the
seems that the assumption of many authors that a
population being studied. Mayhew et al.[17] reported
30% 1RM load maximises power output remains
problematic. Investigating the power-load spectrum that 12 weeks of weight training increased power at
using dynamic (isoinertial) multiarticular motion a fixed absolute load (Pmax increased). Presumably,
would appear to have greater practical significance as the athletes became stronger, the absolute load
to strength and conditioning practitioners and sport became lighter and consequently could be lifted
scientists alike. with greater speed. Thus, the increase in Pmax from
40% to 50% 1RM was due to a 10% increase in
maximal strength. However, this does not seem the
2.1 Upper Body
case when relative loads are used.
The upper-body mean and peak power outputs Baker et al.[15] found that the percentage 1RM
associated with a spectrum of loads can be observed that maximised power output was significantly low-

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
 2005 Adis Data Information BV. All rights reserved.

Maximal Power Training and Improving Athletic Performance


Table I. Loads that maximised mean and peak power output for the upper body

Study Subjects Power measure Maximum power output Maximum power output
(% 1RM or load) (W) [mean ± SD]
Mean power output

Baker[18] 22 M professional rugby league players Concentric-only BP throw across loads 40, 50, NRL: 70kg (51% 1RM) NRL: 600 ± 83
(NRL) 60, 70 and 80kg SRL: 60kg (55% 1RM) SRL: 502 ± 78
27 M college-aged players (SRL)

Baker et al.[15] 31 professional and semi-professional Concentric-only BP throw across loads 40, 50, 55% 1RM 598 ± 99
rugby league players 60, 70 and 80kg 46–62% 1RMa

Cronin et al.[16] 27 M club rugby players Concentric and rebound BP and concentric and 50–70% 1RM 211–356
rebound BP throws across loads of 30, 40, 50,
60, 70 and 80% 1RM

Izquierdo et al.[19] 26 middle-aged (mean age 42y) and 21 Concentric only and stretch-shorten cycle BP 30–45% 1RM for both 237–293
elderly M (mean age 65y) across loads of 0, 30, 45, 60 and 70% 1RM age groups

Izquierdo et al.[20] 70 M weightlifters, middle-distance Concentric only BP across loads of 30, 40, 50, 30–45% 1RM 200–391
runners, handball players, cyclists and 60, 70, 80, 90 and 100% 1RM 2.82–4.86 W/kg
controls

Newton and Wilson[21] 45 M with at least 6mo bench-press Rebound BP throws across loads of 10, 20, 30, 30–40% 1RM
training experience 40, 50, 60, 70, 80, 90 and 100% 1RM

Newton et al.[9] 17 M exercise science students with Concentric only and rebound BP throws across 30–45% 1RM 560–563
6mo weight training experience loads of 15, 30, 45, 60, 75 and 90% 1RM

Peak Power Output

Bemben et al.[22] 31 M college students Rebound BP across loads of 30, 40, 50, 60, 70 50% 1RM
and 80% 1RM

Cronin et al.[16] 27 M club rugby players Concentric and rebound BP and concentric and 50–60% 1RM 463–626
rebound BP throws across loads of 30, 40, 50, 40–70% 1RMa
60, 70 and 80% 1RM

Mayhew et al.[17] 21 M college students Rebound BP across loads of 30–80% 1RM 40% 1RM pre-
intervention
50% 1RM after subjects
increased strength
Sports Med 2005; 35 (3)

Siegel et al.[23] 25 M college students BP across loads of 30, 40, 50, 60, 70, 80 and 40–60% 1RM ~500b
90% 1RM

a Similarly effective to loads that maximised power output.

b Extrapolated from graph.

1RM = one repetition maximum; BP = bench press; M = male; NRL = national rugby league; SRL = student rugby league.

217
 2005 Adis Data Information BV. All rights reserved.

218
Table II. Loads that maximised mean and peak power output for the lower body

Study Subjects Power measure Maximum power output Maximum power output
(% 1RM) (W) [mean ± SD]
Mean power output
Baker et al.[24] 32 professional and semi-professional JS across loads of 40, 60, 80 and 100kg – 55–59% 1RM 1851 ± 210
rugby league players system mass 47–63% 1RMa
Izquierdo et al.[19] 26 middle-aged M (mean age 42y) and Concentric only and stretch-shorten cycle half- 60–70% 1RM for both 391–486
21 elderly M (mean age 65y) squats across loads of 0, 30, 45, 60 and 70% age groups
1RM
Izquierdo et al.[20] 70 M subjects – weightlifters, middle- Concentric only half-squats across loads of 30, 45–60% 1RM 385–755
distance runners, handball players, 40, 50, 60, 70, 80, 90 and 100% 1RM 5.5–9.43 W/kg
cyclists and controls
Sleivert and 30 M rugby, rugby league and SS and TS across loads of 30, 40, 50, 60 and SS: 30–60% 1RM 7.32 ± 1.34 W/kg
Taingahue[25] basketball players 70% 1RM TS: 30–60% 1RM 7.07 ± 1.25 W/kg
Weiss et al.[26] 31 M fitness-trained lifters Concentric-only parallel squats across loads of 30% 1RM 1011 ± 100
30, 60 and 90% 1RM

Peak power output


Bourque and Sleivert[27] 16 males (eight power [six volleyball, Parallel concentric JS across loads of 0, 30, 40, Mean: 14% 1RM Mean: 5216 ± 1234
two badminton], eight endurance 50, 60, 70% 1RM Mode for power Power: 6117 ± 867
athletes) Body mass included athletes 0% 1RM; Endurance: 4315 ± 808
mode for endurance
athletes 30% 1RM
Esliger and Sleivert[28] 21 (11 M and 10 F) volleyball and Parallel concentric JS across loads of 30, 40, 63% 1RM 1766 ± 479
basketball players 50, 60, 70 and 80% 1RM
Siegel et al.[23] 25 M college-aged students Squats across loads of 30, 40, 50, 60, 70, 80 50–70% 1RM ~950b
and 90% 1RM
Sleivert and 30 M rugby, rugby league and SS and TS across loads of 30, 40, 50, 60 and SS: 30–60% 1RM 17.10 ± 3.15 W/kg
Taingahue[25] basketball players 70% 1RM TS: 50–70% 1RM 17.58 ± 2.85 W/kg
Stone et al.[29] 10 subjects with a range of training JS and CJ across loads of 10, 20, 30, 40, 50, Weakest subjects: 10% JS: 3482 ± 443
experience (7wk to >15y) 60, 70, 80, 90 and 100% 1RM 1RM CJ: 3785 ± 376
Strongest subjects: JS: 5635 ± 2577
40% 1RM CJ: 5391 ± 2566
Thomas et al.[30] 19 untrained F Double leg-press 56–78% 1RM 404 ± 22
Sports Med 2005; 35 (3)

Cronin & Sleivert


Weiss et al. [26]
31 M fitness-trained lifters Concentric only squats (parallel) 60% 1RM 1711 ± 188

a Similarly effective to loads that maximised power output.


b Extrapolated from graph.
1RM = one repetition maximum; CJ = countermovement jump; F = females; JS = jump-squats; M = males; SS = split jump-squats; TS = traditional jump-squats.
Maximal Power Training and Improving Athletic Performance 219

er in professional compared with state and city- calculation and, consistent with the results of Bour-
league college-aged rugby league players (see table que and Sleivert,[27] their calculated mean power
I). Reanalysed data from Baker’s previous research values are much higher than other mean power
confirmed this finding, the strongest national league values reported in the literature. The relative load at
players used significantly lower resistances of 47% which they reported mean power to be maximal was,
1RM compared with the 54% 1RM of less strong however, much higher than that reported in Bourque
national league players.[31] It would seem as athletes and Sleivert’s study and consistent with the other
become stronger they can produce greater power literature. Nevertheless, it should be noted that Bak-
outputs with any absolute load, but the ability to er et al.[24] did not measure power output at loads
produce power at a given percentage of their 1RM lighter than 40kg on the bar, so it seems likely that
remains similar as relative resistances increase pro- mean power could have been higher at lighter loads
portionally to maximum strength levels. using this method of calculation.
Other studies have used different calculations to
2.2 Lower Body study the load-power relationship. Sleivert and Ta-
ingahue[25] calculated net power as force exerted
In terms of the lower body, the mean and peak
into the bar by the subject: (bar weight + [bar mass ×
power outputs associated with a range of loads can
acceleration by subject]) × velocity of bar. It appears
be observed in table II. The power outputs are
that Izquierdo et al.[20] and Weiss et al.[26] calculated
reported for a greater variety of movements com-
power output by simply using bar mass relative to
pared with the upper body, and female power out-
1RM. The variety of methods used to calculate the
puts are also represented. The research of Thomas et
power output makes inter-individual comparisons
al.[30] reported a higher Pmax (56–78% 1RM) than
between studies difficult. Clearly a standard method
most other research, which could be attributed to the
for calculating power in resistance training move-
subject’s untrained status, female sex or the differ-
ments needs to be agreed upon. In the meantime,
ent movement used (double leg press). However, in
researchers and practitioners should be aware of the
terms of the squat and its derivatives, no clear trends
implications resulting from including or excluding
are observable between training status, sex, age or
body mass in power calculations for exercises oc-
type of exercise used. Most studies report a
curring in the vertical plane. It seems reasonable to
‘bandwidth’ of loads that maximise power output.
include body mass as part of the resistance athletes
Also, unlike in upper-body exercises, peak power
are working against for exercises occurring in the
has been reported in some instances to occur with no
extra load on the bar (unweighted) or only a light vertical plane. If body mass is not taken into account
load (10–20% 1RM).[27] Bourque and Sleivert[27] when calculating the load for subjects/athletes, inap-
recently reported power results much higher and at propriately heavy loads may be selected for jump-
lighter relative loads than other studies in the litera- squats. For upper-body exercises, only a fraction of
ture that have reported lower peak power to occur at body mass is propelled, therefore, the load-power
heavier relative loads. This is because body mass relation is potentially different and mean and peak
has been included as part of the resistance the ath- power are likely to occur at higher relative loads.
letes are propelling in the jump-squat. When using Besides incorporating body mass into the power
ballistic motion such as jump-squats, it is thought calculations, there are two other reasons why both
appropriate to use system mass to calculate loading Baker et al.[24] and Bourque and Sleivert[27] reported
intensity, as the subject must propel themselves as high power outputs in trained power type athletes.
well as the bar. Excluding body mass from the First, jump-squats were used in both these studies,
equation decreases the total mass component of and ballistic motion of this type has been shown to
force, therefore, decreasing total power output. Bak- increase power output compared with traditional
er et al.[24] also included body mass in their power weight-training movements as used by Izquierdo et

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
220 Cronin & Sleivert

al.[20] and Weiss et al.[26] where there is no projection detail the loading parameters specific to each indi-
of oneself or the bar.[16,32] Secondly, specific neuro- vidual exercise and that this be based on 1RM
muscular adaptations (higher maximal strength or assessment similar to the range of motion being
maximum velocity of contraction) may have influ- prescribed for power training. Qualifying this state-
enced Pmax. In the recent dissertation of Bourque ment, prescription of the load that optimises Pmax
and Sleivert,[27] the major difference in jump-squat needs to be determined for each individual per exer-
power between power and endurance athletes was cise. Research reports the mean response (% 1RM =
the ability of the power athletes to increase the Pmax) for the population being studied, but the
velocity of movement at light loads. Endurance ath- ‘bandwidth’ approach to reporting Pmax adopted by
letes could not increase the velocity of movement most research suggests that there is a range of loads
and, therefore, power decreased markedly as load that maximise power output or more likely that there
decreased. are large inter-individual differences in Pmax.
As stated earlier in this section, the method of
calculating Pmax has a large influence on both abso- 3. Power and Performance:
lute power values and the power-load relationship. Cross-Sectional Research
Additionally, the relative load that maximises peak
Theoretically, the best improvements in athletic
power appears higher than the load that maximises
tasks that involve significant power output (jump-
mean power. Furthermore, it seems that these fac-
ing, sprint, agility and lunge performance) would be
tors may be influenced by the type of exercise
gained by training at the load that maximised an
performed.[25,30] The previously discussed results of individual’s power output using an exercise similar
Edgerton et al.[11] (see section 1) would support such to their athletic activity. However, this presumes
a contention given that Pmax occurred at different that power is the best predictor of athletic perform-
isometric forces depending whether the knee exten- ance and, therefore, it is training to improve power
sors (45%), knee flexors (59%), plantar flexors output that will best facilitate improved perform-
(35%) or dorsiflexors (53%) were used. Therefore, ance. Such an assumption may be misplaced princi-
different exercises (muscle groups) may conceiva- pally because of the diversity of strength/power
bly have differential power outputs. However, measures, flawed methodologies and misrepresent-
Sleivert and Taingahue[25] explain the differences in ed research findings. For example, literature dealing
maximal peak power between the split and tradition- with the development of power tends to mix termi-
al jump-squats reported in their study as being a nology using such terms as power, rate of force
result of the different starting position of both exer- development (RFD), explosive strength (maximum
cises. That is, the 1RM for the low start position of force/time to achieve maximum force) and/or im-
the traditional squat (149.5 ± 22.6kg) was signifi- pulse interchangeably. One needs only to read some
cantly less than for the split squat (206.6 ± 34.4kg). reviews in this area to observe the confusion, power
As a result of the low start position of the traditional has been associated with the ability to exert great
squat, the load was difficult to move initially, but force in a short amount of time (impulse)[33] and
comparable velocities to the split squat were confused with explosive strength or rate of force
achieved later in the lift. Through the range of development.[34] Newton and Kraemer,[35] in consid-
movement in which peak power occurred, both ering methods to increase muscular power, devote
squats had similar bar velocities and absolute loads, much of their discussion to the importance and
although relative loads were very different. There- development of RFD. Sapega and Drillings,[36] in a
fore, traditional squat peak power was maximal at a discussion of the confusion that abounds concerning
higher relative load to that of the split squat. It was the measurement of power, detail how one group of
concluded that the prescription of maximal power authors have calculated peak power by dividing
training using different exercises and ranges should peak torque by the duration of the contraction and

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
Maximal Power Training and Improving Athletic Performance 221

Table III. Intercorrelation matrix between traditional strength and power measures and Zatsiorsky’s measures of explosive strength[40]a
IES RC SG AG MP PP MF PF I100 MS
IES *
RC 0.84 *
SG 0.85 0.68 *
AG 0.80 0.76 0.55 *
MP 0.75 0.43 0.76 0.54 *
PP 0.74 0.45 0.73 0.60 0.99 *
MF 0.80 0.55 0.62 0.52 0.88 0.85 *
PF 0.86 0.61 0.69 0.63 0.92 0.90 0.99 *
I100 0.80 0.54 0.53 0.51 0.89 0.86 0.99 0.99 *
MS 0.81 0.56 0.69 0.57 0.85 0.83 0.94 0.94 0.94 *
a All correlations are statistically significant at p ≤ 0.05.
AG = acceleration gradient; I100 = impulse at 100ms; IES = index of explosive strength; MF = mean force; MP = mean power; MS =
maximal strength; PF = peak force; PP = peak power; RC = reactivity coefficient; SG = starting gradient; * indicates 100% perfect
correlation.

two other studies have used initial RFD as measures activity coefficient, start-gradient and acceleration-
of power.[36] gradient)[39] can be observed from table III. The
relationship of these measures of explosive strength
Misused terminology and misrepresented re-
to more traditional measures of force and power
search findings are also prevalent in the literature. In
suggest that the measures of explosive strength for
a study of 8–12RM weight training performed at
the most part measure different strength qualities
different execution speeds, Young and Bilby[37] de-
than more traditional measures. In particular, the
scribed their two power measures as maximum RFD
reactivity coefficient (r = 0.43–0.61) and A-gradient
and jump height. Neither measure is representative
(r = 0.52–0.63) have less of their variance explained
of power output. Indeed, leg power is generally not
by force and power.
strongly correlated to jump height and there are
limitations in extrapolating results from functional Progressing this contention, it may be that the
tests such as vertical jump to reflect power. Addi- predilection of research and conditioning practice
tionally, some research findings have been inter- on improving power may be misplaced. That is,
preted erroneously and, as such, conclusions about strength qualities such as impulse, RFD or explosive
power development are questionable. For example, strength may better predict athletic performance and
the work of Schmidtbleicher and Buehrle[38] has hence it is the development of these qualities that
been used as justification for the use of high loads research and strength training should focus on. Fur-
for the development of power.[9,15] However, this ther confounding the understanding of power as-
study only measured the changes in maximum force sessment and development is the practical signifi-
and RFD, and changes in these strength qualities are cance of mean and peak power output. The impor-
not necessarily representative of changes in power. tance of these two variables, their relation to
Clearly, in each of these cases, different strength different athletic activities and their development is
capabilities are being investigated or discussed, each not well documented and for the most part poorly
representative of different regions of the force-time understood. The first part of this section will critique
curve as opposed to power the product of force and the literature that has investigated the relationship
velocity. Each of these strength qualities conceiva- between power and athletic performance. Thereaf-
bly needs to be developed in a different manner to ter, the relationship between other strength capabili-
power. For example, the relationship between tradi- ties and athletic performance will be discussed in an
tional measures of force and power, and measures of effort to clarify the importance of Pmax in improving
explosive strength (index of explosive strength, re- athletic performance.

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
222 Cronin & Sleivert

3.1 Upper Body equally effective in explaining the shared variance


between strength/power measures and the perform-
There is a paucity of data concerning the relation- ance measure. However, it should be noted that Pmax
ship between power measures and upper-body func- was not determined in this study but rather the
tional performance. This may be a result of the power outputs associated with a concentric-only
difficulty of finding a suitable upper-body test that is bench-press throw (10kg) were calculated.
as widely accepted as sprinting or jumping. Howev-
er, tasks such as throwing shot-puts, balls or unload- 3.2 Lower Body
ed barbells have been used in the literature. For
example, Mayhew and colleagues[41-43] have ex- In terms of the lower body, the manner in which
amined the relationship between bench-press power power has been calculated and the variety of exer-
and shot-put throw in a number of studies. They cises used to assess power make comparisons be-
found that bench-press power output using an abso- tween studies difficult (see table IV). As a result, the
lute load of 20kg (~40% 1RM) was non-significant- importance of power as a determinant of athletic
ly correlated (r = 0.38) to seated shot-put throw for performance is difficult to disentangle. One ap-
64 female college athletes.[42] For 40 college football proach to clarify the role of power as a predictor of
players, Mayhew et al.[41] found that the seated shot- athletic performance is to discuss power assessment
put throw was significantly correlated to bench- within certain frameworks noting thereafter whether
press absolute power (r = 0.51) and relative power the literature supports such a contention.
output (r = 0.66) assessed using a load of 60% 1RM.
Mayhew et al.[43] also reported that none of the 3.2.1 Cyclic versus Acyclic Assessment
changes in seated shot-put distance were significant- Theoretically, tests of power output that are cyc-
ly correlated to increases in bench-press power out- lic in nature, involve the stretch-shorten cycle (SSC)
put (loads 30–80% 1RM) at the conclusion of a and include horizontal as well as vertical motion
12-week training study. It would seem that the should better predict performance in tasks such as
strength of the relationship between power output running and agility performance. The Margaria-
and seated shot-put throw could possibly be influ- Kalamen step test is one of the more widely used
enced by the mass of the bench-press load, the mass tests used to calculate anaerobic power: power =
of the shot-put, and the strength and sex of the (subject mass [kg] × 9.8 N/kg × 1.02 [vertical dis-
subjects. tance between test stairs])/time taken to ascend test
Baker[31] found that Pmax was significantly relat- stairs. Significant correlations (r = –0.43 to –0.71)
ed (r = 0.46) to an incline bench-press throw (20kg), were reported between power and sprint and agility
an exercise deemed to indicate upper-body speed performance using this test, only when power output
capabilities. Pmax explained only 21% of the vari- was expressed relative to mass (see Mayhew et
ance associated with the so-called functional per- al.,[42] table II). The highest correlation was reported
formance measure. It also should be noted that pow- between the 36.5m (40-yard) dash time and the step
er outputs associated with absolute loads of 40 and test. This makes sense as theoretically the greater the
60kg were significantly related to the performance approach velocity during the test, the greater the
measures (r = 0.42–0.50). Cronin and Owen[44] in- vertical velocity and, therefore, power output. How-
vestigated whether chest-pass distance was related ever, there is still a great deal of unexplained vari-
to various strength and power measures of the upper ance (50–81%) between the step test and the mea-
body as measured by a bench-press throw (10kg) on sures of performance. Some studies have reported
a Smith machine. A significant relationship between that approach velocity and power output are only
maximal strength (r = 0.71), mean power (r = 0.77), moderately correlated[50] and, therefore, it has been
peak power (r = 0.80) and impulse (r = 0.81) and the speculated that something more than vertical veloc-
chest pass were reported. It seems that impulse is ity accounts for power production. The shared vari-

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
 2005 Adis Data Information BV. All rights reserved.

Maximal Power Training and Improving Athletic Performance


Table IV. Relationship between measures of power and performance for the lower body

Study Subjects Power measure Performance measure r


Baker and Nance[45] 20 professional rugby Loaded jump-squats across loads of 10m sprint time –0.02 to –0.08 (NS)
league players 40, 60, 80 and 100kg 40m sprint time –0.52 to –0.61*
Absolute (W) –0.02 to –0.17 (NS)
Relative (W/kg) –0.52 to –0.76*

Chelly and Denis[46] 11 M handball players Average hopping power Maximal track running velocity 0.66*
Average treadmill power (W/kg) 0.20 (NS)
Average treadmill power 0.73*

Driss et al.[47] 18 M volleyball players Anaerobic power (Pmax) on cycle Vertical Jump 0.75**
(6 sec) expressed as W/kg

Kukolj et al.[48] 24 well conditioned M PE Average leg power per kg of body 0–15m time 0.03 (NS)
students mass from continuous jumping 15–30m time 0.26 (NS)
protocol according to Bosco et al.[49]

Mayhew et al.[50] 53 football players Anaerobic power (W) 9m (10yd) dash time 0.16 (NS)
Anaerobic power (W/kg) 36.5m (40yd) dash time –065*
Margaria-Kalamen step test Agility – modified Missouri State 0.21 (NS)
agility run –071*
0.22 (NS)
–0.43*

Meckel et al.[51] 20 F track athletes Wingate anaerobic test (30 sec): 100m sprint from standing start –0.88
10 recreationally trained F maximum power (W/kg)

Nesser et al.[52] 20 sportsmen Wingate 10 sec anaerobic test: 40m sprint time –0.46*
anaerobic power (W/kg)

Sleivert and 30 M rugby league, rugby, Split squat 5m time MP: –0.68**
Taingahue[25] and basketball players Traditional squat 5m time PP: –0.65**
MP: –0.64**
PP: –0.66**

Thomas et al.[30] 19 untrained F Double leg-press peak power output Vertical jump 0.73 (p < 0.004)
Sprint 36.5m (40yd) 0.14 (p < 0.573)
Sports Med 2005; 35 (3)

Young et al.[53] 18 footballers CMJ (power) 20m straight sprint 0.66 (NS)
4 × 20m sprints bouncing football or
changing direction (agility)

CMJ = counter-movement jump; F = females; M = males; MP = mean power; NS = not significant; PE = physical education; Pmax = maximal power output; PP = peak power; * p <
0.05; ** p < 0.001.

223
224 Cronin & Sleivert

ance between the power and performance measures Nance[45] with the absolute power values being the
would certainly support such a contention. only measures significantly related to maximal ve-
Other studies that have used cyclic SSC assess- locity in a 40m sprint. Both studies investigated the
ment include Chelly and Denis[46] who used a tread- relationship between power output and sprint times
mill sprint test and Kukolj et al.[48] who used contin- of athletes over fairly similar distances, the differ-
uous jumping protocol. The best single predictor of ences in dynamometry (jump-squats vs treadmill
maximal running velocity was treadmill forward leg sprints) being the major difference between studies.
power (r = 0.73) in Chelly’s study; however, this It should be pointed out that in the study of Chelley
only accounted for 53% of the variance. Kukolj et and Denis,[46] although relative power was not corre-
al.[48] found no significant relationship between their lated to maximal running velocity, it was strongly
power measures and sprint times. correlated to sprint acceleration (r = 0.80). Of the
Another test of anaerobic power that is cyclic in other studies that expressed power output relative to
nature but does not involve SSC motion is the Win- body mass, the results were mixed with high,[51]
gate cycle test. Three studies[47,51,52] used the power moderate,[47] low[52] or non-significant relation-
output from the Wingate test to predict jump and ships[48] to measures of performance reported. There
sprint performance. Once more the diversity in test- appears to be no clear consensus as to the impor-
ing procedures is apparent, even when the same test tance of normalising power output to body mass for
is used. All three studies calculated power output predicting performance. Also, most normalised
over different time periods. Furthermore Driss et power outputs (the exception being Meckel et al.[51])
al.[47] determined Pmax and used this measure for explained <57% of the common variance associated
comparisons, Nesser et al.[52] used peak power dur- with the performance measure. Meckel et al.[51]
ing their 10-second work bout and Meckel et al.[51] studied a very large group (n = 30) of mixed sprint
used the highest work performed during any 5-sec- ability (mean 100m time ranged from 11.1 seconds
ond work interval. The variety in outcome measures in the fastest group [n = 10] to 14.2 seconds in the
between studies also makes comparisons difficult slowest group [n = 10]). Heterogeneity of subjects
and it is no wonder a wide range of correlations may have contributed to the high correlations report-
between power and performance measures were re- ed.
ported (r = –0.46 to –0.88). From the studies repre-
sented in table IV it appears that the power outputs 3.2.3 Maximum Power versus Power Output
associated with cyclic type motion do not predict The low shared variance between power output
performance any better than acyclic measures. and performance measures reported previously may
be due to the fact that most of the studies have not
3.2.2 Absolute versus Relative Power Output determined Pmax for the assessment that is predict-
It may be that power needs to be normalised to ing the performance measure. Baker and Nance[45]
body mass for power output to be truly representa- reported that Pmax/kg was significantly related to
tive of the power needed to run and jump. Once 10m (r = –0.56) and 40m (r = –0.76) sprint perform-
more, the literature is confusing. Two studies[45,46] ance. However, the strength of these correlations
have represented their power outputs in absolute were very similar to the correlations associated with
(W) and relative (W/kg) terms. The absolute power other loads and in the case of 10m sprint perform-
outputs of Baker and Nance[45] were not significant- ance, the correlations between Pmax and speed were
ly related to 10 or 40m sprint performance; however, lower than the correlations reported for some of the
when expressed relative to body mass, significant other loads. Driss et al.[47] also determined Pmax for a
relationships were reported (r = –0.52 to –0.76), 6-second sprint test on a cycle at different braking
likely because body mass must be accelerated in forces. The braking force that maximised Pmax was
sprinting. However, some of the findings of Chelly recorded and related to functional performance (ver-
and Denis[46] are the antithesis of Baker and tical jump). Unlike Baker and Nance,[45] the other

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
Maximal Power Training and Improving Athletic Performance 225

power outputs associated with various loads and measured. Reviewing research that has investigated
subsequent relationships were not reported in this the ability of a variety of strength and power mea-
study, therefore, conclusions cannot be made as to sures to predict athletic performance, may give bet-
the significance of the load that optimises Pmax in ter insight into the importance of power as a deter-
relation to other loads. In both studies, however, minant of athletic success.
Pmax accounted for <58% of the shared variance Cronin et al.[40] assessed the strength, power and
between the performance measures. explosive strength of the leg musculature on a su-
It appears that none of these frameworks clarify pine squat and related these measures to lunge per-
the importance of power output or Pmax as predic- formance (lunge foot contact times: 0.354 sec +
tors of performance more so than the other. Given 0.063). The only significant strength predictors of
the great variety of assessment techniques, it might lunge performance were all explosive strength mea-
have been expected that one type of assessment may sures (index of explosive strength [r = 0.62], reactiv-
predict performance to better effect. This was not ity coefficient [r = 0.61], starting gradient [r = 0.69]
the case, the greatest shared variance between a and acceleration gradient [r = 0.59]). Maximal
measure of power and performance being 77% and strength, mean and peak power were not significant-
most coefficients of determinations approximately ly related to performance of this movement.[40] In
50% or less. It may be that power is not an important another study, Sleivert and Taingahue[25] investigat-
determinant of performance and other strength mea- ed the relationship between sprint start performance
sures may be of equal or greater importance or (5m sprint time) and strength and power variables
power measurements have not be appropriate for the determined from concentric jump-squats in 30 male
performance. athletes. Both average and peak power expressed
relative to body mass were significantly related to
It has been suggested that for activities that re-
5m sprint time (r = –0.64 to –0.68). Force (r = 0.59)
quire fast force production (100–300ms) such as
and bar velocity (r = 0.40) were also significantly
throwing, jumping and sprinting,[6] that the rate at
related to 5m sprint time. However, with the excep-
which force is developed is the most important
tion of peak force, these relationships were much
determinant of athletic success.[54] Wilson et al.[55]
less substantial than those demonstrated for average
investigated the relationship of a series of isometric,
and peak power. During the sprint start, the body
concentric and SSC rate of force development tests
had to be accelerated rapidly from stationary and the
performed in an upright squat position to sprinting
propulsive impulse was large, so there may have
performance. Of the 20 force-time variables gener-
been a greater reliance on high force production as
ated using a modified Smith machine over a force
opposed to high movement velocity.
platform, the concentric force at 30ms was the only
measure significantly correlated to sprint perform- Young et al.[56] using a similar methodology to
ance (r = –0.616) and, in addition to concentric Wilson et al.[55] investigated the relationship be-
maximum RFD, were the only measures able to tween 27 strength and power measures and the
effectively discriminate between good and poor per- sprinting performance of 20 elite junior track and
formers.[55] The authors emphasise the superiority of field athletes (11 males and 9 females). There was
concentric RFD tests over and above isometric and no mention whether the pooling of the male and
SSC RFD tests and suggest their inclusion in sport female subjects were investigated for bipolar distri-
science test batteries. It should be remembered, bution (by sex), which can result in artificially high
however, that concentric RFD and force at 30ms correlations. Therefore, the magnitude of the corre-
explained <38% of the variance associated with 30m lations need to be interpreted with caution. Young
sprint performance. Furthermore, this type of meth- and colleagues[56] assessed vertical jumping move-
odology does not clarify whether power is a better ments utilising purely concentric, SSC and isometric
predictor of athletic performance, as power was not contractions performed over a force platform, found

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
226 Cronin & Sleivert

that concentric strength measures were the best into all the mechanisms responsible for performance
predictors of sprint performance. The best predictors of a task. Secondly, other factors such as strength
of starting performance (time to 2.5m) were all measures, body mass, flexibility and leg length will
obtained from a concentric-only jump-squat and have diverse effects on the statistical models. Based
included three force measures (maximum dynamic on these results, it is suggested that the sports train-
strength/weight [r = –0.86], force in 100 ms/weight er, sport scientist or clinician should not rely solely
[r = –0.73], maximum force [r = –0.72] and one on a single power measurement to predict perform-
power measure (average power/weight [r = –0.74]). ance or readiness to return to activity after injury.
The single best predictors of maximum sprinting Rather, research needs to determine the influence of
speed were the force relative to bodyweight generat- these other factors on athletic performance. It may
ed after 100ms from the start of the concentric jump be that several factors in combination with power
movement (r = –0.80) and maximum force (r = measures will provide the best predictive capabili-
–0.79). Average power relative to bodyweight was ties of functional performance. Therefore, the chal-
also strongly related to maximum velocity (r = lenge is to develop assessment batteries that provide
–0.79). Together, the findings of these studies sug- insights into the key mechanisms responsible for the
gest that for many types of activities, various rates of performance of a task. It must also be remembered
force development and maximal force measures that a significant relationship to performance does
may be as predictive of performance as measures of not imply that a particular strength or power attri-
maximal power. The magnitude of these relation- bute, when loaded during training, will enhance
ships appear to depend upon the kinetics and kine- performance. Only longitudinal data from training
matics of the performance in question, and specifici- studies can provide information regarding optimal
ty of neuromuscular demand appears to be a critical training methods, if indeed optimal methods exist.
factor in determining the strength or power attribute
that best predicts performance. It should be remem- 4. Power and Performance:
bered, however, that just because a strength or pow- Training Studies
er attribute is related to performance, that does not
necessarily indicate that training that particular attri- Research that has investigated the development
bute will enhance performance. of power is typified by a great deal of variation in
the methodological approaches used. The scope of
3.2.4 Summary this variation makes comparisons difficult and
Implicit in the small number of meaningful corre- hence definitive conclusions practically impossible.
lations and large amount of unexplained variance in For example, the vast majority of research has been
the data of table IV is that sport scientists need to relatively short in duration (8–12 weeks) and, there-
formulate better methods, models and theories to fore, the application of findings to long-term train-
contribute significantly to knowledge that is useful ing is questionable as the influence of neural and
to athletes and their coaches in terms of power and morphological mechanisms change with training
performance. It has been suggested that sport scien- duration.[57] Research in this area is also typified by
tists adopt a specific functional term to denote the a wide spectrum of loading parameters that include
ability to perform high-velocity explosive force- differences in:
generating movements and limit the term ‘power’ to • volume
its correct usage as defined by Newtonian mechan- • intensity (% 1RM)
ics.[36] Whether this is necessary is debatable. It may • total work output
be that the preoccupation of correlational studies to • tempo of concentric-eccentric contractions
find the best power predictors of functional perform- • frequency
ance is fundamentally flawed. First, one power mea- • rest/recovery time – density
sure cannot adequately express or provide insight • type of contractions.

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
Maximal Power Training and Improving Athletic Performance 227

Further confounding our understanding in this with plyometric training on a variety of performance
area are the modes of dynamometry used and variety measures. Conversely Komi et al.[61] compared the
of strength power measures reported. These issues effect of heavy-load training versus light-load train-
have been discussed in section 3. Finally, many ing combined with explosive jump training on vari-
different muscle groups have been studied, limiting ous performance measures. Any performance
the ability to generalise results, especially in the case changes, however, cannot be attributed to a load
of uni- and bi-articular muscles. To discuss each of effect as one treatment group has incorporated addi-
these limitations is outside the scope of this article. tional plyometric training. Unfortunately this type of
However, a brief mention will be made of some methodology is not uncommon and as such does not
issues that are thought influential to subsequent dis- offer insight into specific loading intensities that
cussion. affect power and performance.
First, a clear understanding of each research Further confounding our understanding of the
methodology needs to be realised, so as the interpre- effects of resistance training on the development of
tation and application of their findings are not mis- power is the preponderance of research using novice
represented. For example, the study of Schmidt- weight trainers or students as subjects (see table V).
bleicher and Buerhle[38] compared the effects of This is due to the accessibility of such subjects to
three types of training regimes (maximum load researchers. Novice subjects and/or student popula-
[90–100% maximal voluntary contraction, MVC], tions are generally easy to access and it is easy to
power load [45% MVC] and hypertrophy load [70% have a control group in such designs. However, it is
MVC]) on various neuromechanical and morpho- more problematic to find a suitable cohort of sub-
logical variables. In terms of tracking these changes, jects from an athletic population and practically
an isolated isometric elbow extension movement impossible to ask a group of athletes to cease train-
was used. Consequently, power was not assessed as ing and act as controls. Nonetheless, it has been
no movement took place. Furthermore, isometric shown that novices respond in a generic manner to a
and dynamic contractions have been shown to differ very broad range of resistance training stimuli.[62,63]
in terms of their physiology and neuromechanics.[58] Thus, the validity of generalising findings from nov-
Similarly, the findings of Hakkinen et al.[59] are used ice subjects to athletes with experience in weight
to support the use of light-load explosive jump train- training needs to be done so with caution as the
ing for power development. However, there were no findings may in fact be compromised by the
other training or control groups cited in this paper trainability of the novice subjects.
and the changes in performance from the 24 weeks
Volume is commonly expressed as the total prod-
of training were tracked by measuring isometric
uct of repetitions, sets and load (% 1RM). Equating
force and various force-time parameters. These arti-
by volume is the most common method by which
cles, however, are often quoted in power training
research compares the effect of load on various
literature even though power has not been measured.
outcome measures. Alternative methods include
The application of such findings for improving the
equating with total time under tension as measured
power of multi-articular dynamic motion would ap-
by electromyographic (EMG) activity or total
pear problematic on a number of counts.
mechanical work performed.[67] A great deal of re-
Another common methodological problem when search has failed to equate loading between training
studying the effect of load is that authors combine protocols in any form[38,65,69,70] and, as a result, the
training methods, which make the effects of the findings of such research must also be interpreted
independent variable impossible to disentangle. For with caution especially if the effect of load is being
example, Lyttle et al.[60] investigated the effects of investigated. For example, Harris et al.[65] studied
maximal power training (30% 1RM) versus heavy- the effects of different training protocols on well
load training (6–10RM or 75–83% 1RM) combined trained subjects. One group used heavy-load squats

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
228 Cronin & Sleivert

Table V. Studies that have examined the effect of load on power output and performance
Study Subjects Training loads used Power and performance measures
Cronin et al.[64] 21 F provincial netball players (1) 60% 1RM Mean and peak power
with no recent weight-training (2) 80% 1RM Netball throw velocity
experience (3) Control
Equivolume training for 10wk
Harris et al.[65] 51 university football players (1) 30% 1RM Average VJ power
with at least 1y of weight- (2) 80% 1RM Peak VJ power
training experience (3) Com 30% and 80% 1RM Standing long jump
Not equivolume – 13wk MK step test
Agility test
30m sprint
McBride et al.[66] 26 athletic men with 2–4y (1) 30% 1RM Peak power
strength-training experience (2) 80% 1RM 20m sprint time
(3) Control Agility T-test
Equivolume training for 8wk
Moss et al.[67] 31 well trained PE students (1) 15% 1RM Power-load spectrum
(2) 35% 1RM
(3) 90% 1RM
Equivolume training for 9wk
Scmidtbleicher and 30 M PE students (1) 30% MVC No power measures
Haralambie[68] (2) 90–100% MVC Maximal speed of push-off movement
(3) Control
Equivolume training for 8wk
Schmidtbleicher and 59 M students (1) 45% MVC No power measures
Buehrle[38] (2) 70% MVC No performance measures
(3) 90–100% MVC
Not equivolume – 12wk
Wilson et al.[69] 64 subjects with 1y weight-
(1) 30% 1RM No measures of power
training experience (2) 6–10RM = 75–83% 1RM JS Ht
(3) Plyometric CMJ Ht
(4) Control Sprint and cycle tests
Not equivolume – 10wk
1RM = one repetition maximum; CMJ = counter-movement jump; Com = combined training; F = females; Ht = height; JS = jump-squat; M =
males; MK = Margaria-Kalamen step test; MVC = maximal voluntary contraction; PE = physical education; VJ = vertical jump.

of 80–85% 1RM (high force [HF]), one with lighter standing long jump, significantly greater increases
load squats of approximately 30% 1RM (high power in the HP (3.45%) group than COM (1.63%) or HF
[HP]), and one with a combination (COM) of train- (1.29%) groups were observed. These differential
ing methods. Subjects were instructed to perform all training effects are interesting; however, it may be
lifts as explosively as possible although not actually that the noted training effects were not due to the
jump. Pre- and post-training tests included: 1RM different kinematics and kinetics characteristics of
squat, 1RM one-quarter squat, 1RM mid-thigh pull the various training protocols, but rather the differ-
and various tests of speed and jump height. After 9 ent training volumes of each group.
weeks of training it was observed that performance Wilson et al.[69] is an often-quoted study that
changes reflected contraction force specificity. That experiences the same problem. These researchers
is, the COM and HF groups increased 1RM strength examined the effects of three different training
significantly more than the HP group (24.7%, 21.9% modes on strength and power outputs over the
and 9.6% for COM and HF and HP, respectively, course of 10 weeks of training. Previously trained
average increase across squat and one-quarter squat subjects were divided into three subject groups. One
1RM). However, for some low-force, high-velocity trained with traditional ‘heavy’ (6–10RM) squats
measures of functional performance such as the (TR), one with jump-squats at Pmax (around 30% of

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
Maximal Power Training and Improving Athletic Performance 229

maximum isometric force [MP]) and one with ply- listic’ strength training[35] and is thought superior to
ometric unloaded depth jumps (0.2–0.8m [PL]). traditional strength-training methods as the velocity
Tests before and after the training period included and acceleration/deceleration profiles better approx-
sprint time, vertical jump height and peak power in a imate the explosive movements used in athletic per-
6-second cycle test. The results indicated that the formance.[21,32]
MP training group demonstrated the best overall Careful consideration also needs to be given to
improvements in functional performance. Increases the type of isoinertial assessment used in training
in countermovement jumps (17.6%) and jump- studies, as assessments must balance between being
squats (15.2%) were significantly greater than the specific to the functional task whilst being suffi-
TR (4.8% and 6.3%) and PL (10.3% and 6.5%) ciently different from training so that it does not
groups. They concluded after a training study that advantage one of the training groups. For example,
compared heavy weight training (6–10RM), ply- it can reasonably be expected that a group training
ometric training (drop jumps from 0.2–0.8m) and with heavy loads would achieve greater improve-
maximum power training (30% 1RM) that the Pmax ments in 1RM than a group training with lighter
training produced the best overall results.[69] Once loads. Therefore, assessments should either include
more, intensity and volume have not been equated tests across a range of loads or at a ‘neutral’ load that
and, therefore, it is impossible to disentangle the will not bias the results for any particular group.
training effects. That is, the results from such studies
are difficult to interpret as the reported differences Finally, all the papers cited in table V experience
between various training protocols may in fact be a basic but fundamental problem if the effect of
contaminated by differences in training volume, training at Pmax is to be established. No paper has
rather than the specific kinematic and kinetic char- established Pmax for their respective subjects, the
acteristics of the different loading intensities. Mak- training load selected has been based on previous
ing conclusions about the efficacy and/or adapta- research. The perils of such an approach have been
discussed throughout this paper with many findings
tions of various training protocols that are not equat-
such as Edgerton et al.,[11] Faulkner et al.,[12] Kaneko
ed in some manner would appear highly
et al.,[13] Schmidtbleicher and Buehrle,[38] misinter-
questionable.
preted and/or misrepresented. The dangers of such
Further complicating our understanding in this an approach are further reinforced if tables I and II
area is that changes in performance may be related are observed, that is which load (% 1RM) should be
to the movement pattern used rather than differences used as Pmax for a training study. If the effects of
in loading intensity. For example, in the Wilson et Pmax training are to be understood, Pmax needs to be
al.[69] article, it may be that the superior performance determined specific to the population and training
of the maximum power groups was not due to train- exercise used. The power loads selected for study
ing at the ‘optimal load’ but rather that jump-squat have mostly been based on research (e.g. Kaneko et
training (ballistic training) offers greater movement al.[13]) that cites the superiority of lighter loads
pattern specificity than more traditional strength- (30–45% 1RM). Tables I and II and previous discus-
training methods. Traditional strength-training tech- sions of research such as that of Kaneko et al.[13]
niques in which a bar is held at the completion of the would suggest that this is not advisable practice.
motion have been criticised due to large decelera- Only one study[64] based their selection of Pmax on
tions during the concentric phase, proportional to their own previous research using similar athletes
the load and, therefore, velocity of movement.[32,71] and assessment equipment to their original re-
Strength training that allows the projection of the search.[16] In this study, the effects of heavy-load
load avoids this problem by allowing the athlete to training (80% 1RM) and maximum power training
accelerate the bar throughout a greater range of (60% 1RM) on the power and chest-pass perform-
movement. Such training has been described as ‘bal- ance of semi-elite netball players were compared.

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
230 Cronin & Sleivert

Significantly greater mean and peak power outputs three groups at loads ≤50% 1RM but for the higher
(40% 1RM bench-press throw) were observed for loads (70% and 90% 1RM) the increase was signifi-
the heavy-load group; however, no significant dif- cantly larger for G90 and G35 compared with G15.
ferences in chest-pass throw velocity between train- In essence, the findings illustrate loads of 35% and
ing groups were found. However, it should be noted 90% 1RM were equally effective in improving pow-
that the players had little or no weight-training expe- er output across a spectrum of loads.
rience. Another study investigating the effects of heavy-
The types of problems described in this section (80% 1RM) versus light-load (30% 1RM) training
are symptomatic of research in this area, the reader on strength, power and speed in experienced weight
having to be discerning in their choice of literature trainers was conducted by McBride et al.[66] Jump-
to shape their knowledge and practice. To gain a true squats were used for both training and testing, and
appreciation of the effect of load on power develop- groups were equated for volume using sets × repeti-
ment, research methodologies need to equate the tions × load (%RM). After 8 weeks of training, both
load lifted in some manner. Ideally, well trained groups increased 1RM significantly (10.17% and
subjects should be trained at different loads (% 8.23% for 80% 1RM and 30% 1RM groups, respec-
1RM), the effects of which need to be reported as tively) with no significant difference between
changes in mean and/or peak power and/or changes groups. Similarly, there were no significant be-
in functional performance. Unfortunately, the num- tween-group differences in peak force, peak power
ber of studies that have adopted such an approach or jump height pre- and post-training. Of the three
are few (see table V) and, as mentioned previously, sprint (5, 10 and 20m sprint time) and agility (T-test
none of these studies determined or trained at Pmax. time) measures, the 30% 1RM training proved supe-
Therefore, the importance of Pmax remains a mys- rior to 80% 1RM training on only one measure (10m
tery. Nonetheless, two studies seem to have resolved sprint time). The results of these two studies suggest
some of the limitations highlighted in previous sec- that there is very little difference in the effects of
tions and may offer insights into optimising the load heavy- and light-load training in terms of power and
selected during resistance training strategies to de- performance.
velop maximal power and enhance performance. The lack of a differential training effect in these
[67]
Moss et al. had three groups of subjects train studies using quite different loads may be explained
the elbow flexors of the non-dominant arm while the if training velocity and actual movement velocity of
dominant arm served as a control. The three groups a task are compared. For example, the velocities
trained at 90% 1RM for two repetitions (G90), at attained during a concentric-only bench-press throw
35% 1RM for seven repetitions (G35) and at 15% and a rebound bench-press throw on a modified
1RM for 15 repetitions (G15), respectively. All Smith-Press machine can be observed from figure
groups trained with three to five sets and were 1.[64] Such motion has been reported to more closely
equated for total time under tension based on EMG simulate the velocity and acceleration profiles asso-
activity (muscle activation). Subjects were en- ciated with throwing.[32] The velocity profiles ob-
couraged to perform each lift as fast as possible, but tained from these 27 male athletes are similar to
the weight was held rather than projected. Measure- those recorded in other research of this kind.[9,32]
ments before and after the study included 1RM and For sports-specific motion that uses muscles sim-
various force and power outputs. Power was tested ilar to the bench press the comparison of actual
across a range of loads from 2.5kg to 90% 1RM and movement velocity to training velocity make for
increased for all loads tested in G90 and G35. An interesting analysis. Ritzdorf[72] details release ve-
increase in a lighter range of loads from 15% to 50% locities of 13 m/sec for the shot-put motion. Average
1RM was observed for G15. The increases in power release velocities of 11.98 m/sec for the chest pass
output were not significantly different between the of semi-elite female netball players have also been

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
Maximal Power Training and Improving Athletic Performance 231

1.1 Concentric bench-press throw


Rebound bench-press throw
1.0

0.9

0.8
Average velocity (m/sec)

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
30 40 50 60 70 80
Percentage 1RM
Fig. 1. Average velocities and standard deviations associated with the concentric bench-press throw and the rebound bench-press throw
over a range of loading intensities (30–80% one repetition maximum [1RM]).

reported.[64] Furthermore, other studies (as reported rapidly develop force. This assertion must be inter-
by Ritzdorf[72]) have reported that the segment ve- preted cautiously as Schmidtbleicher and Buehrle[38]
locities (shoulder, elbow and hand) combined to found the greatest improvements in RFD resulted
produce racquet head velocities of approximately 30 from maximum load training (90–100% MVC) as
m/sec for the tennis serve. It would seem that there opposed to maximum power (45% MVC) training
are clear differences between the velocities associat- (34% and 11% improvements, respectively). It
ed with the most common strength-training loading should be remembered that rate of force develop-
intensities (30–80% 1RM) and the actual movement ment can be independent of external movement ve-
velocity of a sport-specific task. There is no doubt locity. In addition, maximal rate of force develop-
that resistance strength training can improve func- ment and motor unit activation in a maximum effort
tional performance. However, the importance of contraction is relatively constant for an individual
load in terms of velocity specificity would appear and is generally not influenced by the external load
questionable, due to the disparity between the actual or speed of movement.[73] In Schmidtbleicher and
movement velocities of most athletic tasks and the Buehrle’s study[38] both groups were instructed to
training velocities achieved during weight training. develop their contractions as fast as possible. It may
It may be that load is not as important as many think be that irrespective of load and limb velocity, the
and other factors may be more important to better repeated intent to move ‘explosively’ is the impor-
develop and explain improvement to functional per- tant stimulus for muscular adaptation. The findings
formance. For example, as suggested in section 3, it of Behm and Sale[74] support such a notion, as they
may be that rate of force development rather than found that regardless of the actual velocity of move-
power output is a more important determinant of ment (isometric versus isokinetic), it was the inten-
athletic success. Newton and Kraemer[35] stated that tion to execute a high-velocity movement, which
heavy loading (70–120% 1RM) does not improve resulted in high-velocity adaptation and substantial
the maximum rate of force development, rather ex- increases (26%) in rate of force development. Along
ercises such as explosive jump training using resis- these lines, it has been recently shown with surface
tances of 30–60% 1RM best increase the ability to electromyography, that at loads between 30–80%

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
232 Cronin & Sleivert

1RM, power, force and velocity values differ, but ance to better effect and direct assessment and con-
muscular activation is similar during maximal effort ditioning practice in a more efficient and systematic
ballistic jump-squats.[19] Thus, when the intent to fashion.
move is maximal, any differences in training effects If Pmax is found to be important, then each indi-
due to loading differences are likely due to differ- vidual’s Pmax needs to be determined and they then
ences in the mechanical stimuli at the muscle, not train at this load. The predilection of research to
activation differences. train all subjects at one load (e.g. 30% 1RM) is
fundamentally flawed due to inter-individual Pmax
5. Conclusions differences, which may be ascribed to factors such
as training status (strength level) and the exercise
Many problems have been magnified throughout (muscle groups) used. Pmax needs to be constantly
this article that currently exist within this field of monitored and adjusted as research suggests that it is
study. From the use of fundamental definitions and transient. In terms of training studies, experienced
terminology to training studies, the literature ap- subjects should be used, volume equated and the
pears hewn with confusion and methodological outcome measures clearly defined and measured
problems. It is hoped that this article has eliminated (i.e. mean power and/or peak power).
some of this confusion and clarified the type of Until training studies address the limitations dis-
research needed if we are to advance our knowledge cussed throughout this paper, the best and safest
and practice on the importance of maximising power course of action for those interested in improving
output and its transference to athletic performance. the power output of muscle may be to use a mixed
Sport scientists are urged to formulate research de- training strategy using both heavy and light loads.
signs that result in meaningful and practical infor- Realising that all human movement is an integration
mation that assists coaches and strength and condi- of force and velocity, such an approach is intuitively
tioning practitioners in the development of their appealing. That is, most sports involve a mixture of
athletes. activities that span the force-velocity capability of
In summary, it is apparent there is a need for a muscle. For example, the shot-putter has to drive
great deal more research on the importance of train- their quite sizeable mass through the circle before
ing at Pmax and whether it advantages athletic per- throwing a relatively light shot-put. Rugby players
formance over and above other loading intensities. not only have to wrestle and tackle each other but
First, the importance of Pmax to athletic performance also kick and throw a ball. Intuitively, it would seem
needs to be established. It must be remembered that prudent to continuously adjust the resistances used
power is only one aspect that affects performance for power training, as athletic performance is typi-
and it is quite likely that other strength measures fied by many force-velocity characteristics. Further-
may be equally if not more important for determin- more, as one of the principles of training is variation
ing the success of certain tasks. The coach and/or (periodisation) this approach would seem most logi-
sport scientist must be aware of this and identify cal.
those strength qualities that are critical determinants
of their athletic activity and thereafter devise appro- Acknowledgements
priate assessment strategies and training program- No sources of funding were used to assist in the prepara-
mes so as these determinants are improved in a tion of this review. The authors have no conflicts of interest
systematic fashion. It is suggested that instead of that are directly relevant to the content of this review.
using a correlational analysis that a regression ap-
proach be used to find predictor models that include References
other anthropometric and physiological (including 1. Counsilman JE. The importance of speed in exercise. Athletic J
1976; 56 (9): 72-5
other strength qualities) measures. As such, a battery 2. Behm DG. Surgical tubing for sport and velocity specific train-
of two or more tests may predict athletic perform- ing. Nat Strength Cond Assoc J 1988; 10 (4): 66-70

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
Maximal Power Training and Improving Athletic Performance 233

3. Poprawski B. Aspects of strength, power and speed in shot put 23. Siegel JA, Gilders RM, Staron RS, et al. Human muscle power
training. Nat Strength Cond Assoc J 1987; 9 (6): 39-41 output during upper- and lower-body exercises. J Strength
4. Spassov A. Special considerations when programming for Cond Res 2002; 16 (2): 173-8
strength and power for athletes: part 1. Nat Strength Cond 24. Baker D, Nance S, Moore M. The load that maximizes the
Assoc J 1988; 10 (4): 58-61 average mechanical power output jump squats in power trained
5. Verkhoshansky YV, Lazarev VV. Principles of planning speed athletes. J Strength Cond Res 2001; 15 (1): 92-7
and strength/speed endurance training in sports. Nat Strength 25. Sleivert GG, Taingahue M. The relationship between maximal
Cond Assoc J 1989; 11 (2): 58-61 jump-squat power and sprint acceleration in team sport ath-
6. Tidow G. Aspects of strength training in athletics. N Studies letes. Eur J Appl Physiol 2002; 87: 264-71
Athl 1990; 1: 93-110 26. Weiss LW, Fry AC, Magu B, et al. Relative external loads
7. Hill AV. The heat of shortening and the dynamic constants of eliciting maximum concentric force and power during non-
muscle. Proc R Soc Lond 1938; 126: 136-95 countermovement squats. National Strength and Conditioning
8. Josephson RK. Contraction dynamics and power output of Association National Conference XXV; 2002, Las Vegas
skeletal muscle. Annu Rev Physiol 1993; 55: 527-46 27. Bourque S, Sleivert GG. Determinants of load at peak power
9. Newton R, Murphy A, Humphries B, et al. Influence of load and during maximal effort squat jumps in endurance and power
stretch shortening cycle on the kinematics, kinetics and muscle trained athletes [dissertation]. Fredericton: University of New
activation that occurs during explosive upper-body move- Brunswick, 2003
ments. J Appl Physiol 1997; 75: 333-42 28. Esliger DW, Sleivert GG. The neuromechanics of maximal
10. Toji H, Suei K, Kaneko M. Effects of combined training loads effort squat jumps [dissertation]. Fredericton: University of
on relations among force, velocity, and power development. New Brunswick, 2003
Can J Appl Physiol 1997; 22 (4): 328-36
29. Stone MH, O’Bryant HS, McCoy L, et al. Power and maximum
11. Edgerton VR, Roy RR, Gregor RJ, et al. Morphological basis of strength relationships during performance of dynamic and
skeletal muscle power output. In: Jones NL, McCartney N, static weighted jumps. J Strength Cond Res 2003; 17 (1):
McComas AJ, editors. Human muscle power. Champaign (IL): 140-7
Human Kinetics; 1986: 43-63
30. Thomas M, Fiatarone MA, Fielding RA. Leg power in young
12. Faulkner JA, Claflin DR, Cully KK. Power output of fast and
women: relationship to body composition, strength and func-
slow fibres from human skeletal muscles. In: Jones NL, Mc-
tion. Med Sci Sports Exerc 1996; 28 (10): 1321-6
Cartney N, McComas AJ, editors. Human muscle power.
Champaign (IL): Human Kinetics, 1986: 81-93 31. Baker D. A series of studies on the training of high-intensity
13. Kaneko M, Fuchimoto T, Toji H, et al. Training effect of muscle power in rugby league football players. J Strength
different loads on the force-velocity relationship and mechani- Cond Res 2001; 15 (2): 198-209
cal power output in human muscle. Scand J Sports Sci 1983; 5: 32. Newton RU, Kraemer WJ, Hakkinen K, et al. Kinematics,
50-5 kinetics and muscle activation during explosive upper body
14. Mastropaolo JA. A test of the maximum-power stimulus theory movements. J Appl Biomech 1996; 12: 31-43
of strength. Eur J Appl Physiol 1992; 65: 415-20 33. Hedrick A. Literature review: high speed resistance training.
15. Baker D, Nance S, Moore M. The load that maximizes the Nat Strength Cond Assoc J 1993; 15 (6): 22-30
average mechanical power output during explosive bench 34. Haff GG, Potteiger JA. A brief review: explosive exercise and
press throws in highly trained athletes. J Strength Cond Res sports performance. Strength Cond J 2001; 22 (3): 13-20
2001; 15 (1): 20-4 35. Newton R, Kraemer W. Developing explosive muscular power:
16. Cronin JB, McNair PJ, Marshall RN. Developing explosive implications for a mixed methods training strategy. Strength
power: a comparison of technique and training. J Sci Med Cond J 1994; 6: 36-41
Sport 2001; 4 (1): 59-70 36. Sapega AA, Drillings G. The definition and assessment of
17. Mayhew JL, Johns RA, Ware JS, et al. Changes in absolute muscular power. J Orthop Sports Phys Therapy 1983; 5 (1):
upper body power following resistance training in college 7-19
males. J Appl Sport Sci Res 1992; 6 (3): 187
37. Young BW, Bilby EG. The effect of voluntary effort to influ-
18. Baker D. Comparison of upper-body strength and power be- ence speed of contraction on strength, muscular power, and
tween professional and college-aged rugby league players. J hypertrophy development. J Strength Cond Res 1993; 7 (3):
Strength Cond Res 2001; 15 (1): 30-5 172-8
19. Izquierdo M, Ibanez J, Gorostiaga E, et al. Maximal strength
38. Schmidtbleicher D, Buehrle M. Neuronal adaptation and in-
and power characteristics in isometric and dynamic actions of
crease of cross-sectional area studying different strength train-
the upper and lower extremitiesin middle aged and older men.
ing methods. In: Jonsson B, editor. Biomechanics. Champaign
Acta Physiol Scand 1999; 167: 57-68
(IL): Human Kinetics, 1987
20. Izquierdo M, Hakkinen K, Gonzalez-Badillo JJ, et al. Effects of
long-term training specificity on maximal strength and power 39. Zatsiorsky VM. Science and practice of strength training.
of the upper and lower extremities in athletes from different Champaign (IL): Human Kinetics, 1995
sports. Eur J Appl Physiol 2002; 87: 264-71 40. Cronin JB, McNair PJ, Marshall RN. Lunge performance and its
21. Newton R, Wilson G. The kinetics and kinematics of powerful determinants. J Sports Sci 2003; 21: 49-57
upper body movements: the effect of load. In: Bouisset S, 41. Mayhew JL, Bemben MG, Piper FC, et al. Assessing bench
Metrai S, Monod H, editors. XIVth international series on press power in college football players: the seated shot put. J
biomechanics. Champaign (IL): Human Kinetics, 1993: 936-7 Strength Cond Res 1993; 7 (2): 95-100
22. Bemben MG, Rohrs DM, Bemben DA, et al. Effect of resistance 42. Mayhew J, Bemben M, Rohrs D, et al. Specificity among
training on upper body strength, power and performance. J anaerobic power tests in college female athletes. J Strength
Appl Sport Sci Res 1991; 5 (3): 162-71 Cond Res 1994; 8 (1): 43-7

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)
234 Cronin & Sleivert

43. Mayhew JL, Ware JS, Johns RA, et al. Changes in upper body 61. Komi PV, Suominen H, Heikkinen E, et al. Effects of heavy
power following heavy-resistance strength training in college resistance and explosive-type strength training methods on
men. J Sports Med 1997; 18: 516-20 mechanical, functional, and metabolic aspects of performance.
44. Cronin J, Owen G. Upper body strength and power assessment In: Komi PV, editor. Exercise and sport biology. Champaign
in women using a chest pass. J Strength Cond Res 2004; 18 (3): (IL): Human Kinetics Publishers, 1982: 91-102
401-4 62. Chestnut JL, Docherty D. The effects of 4 and 10 repetition
45. Baker D, Nance S. The relation between running speed and maximum weight-training protocols on neuromuscular adapta-
measures of strength and power in professional rugby league tions in untrained men. J Strength Cond Res 1999; 13 (4):
players. J Strength Cond Res 1999; 13 (3): 230-5 353-9
46. Chelly SM, Denis C. Leg power and hopping stiffness: relation- 63. Hakkinen K. Neuromuscular and hormonal adaptations during
ship with sprinting performance. Med Sci Exerc Sports 2001; strength and power training. J Sports Med Phys Fitness 1989;
33 (2): 326-33 29 (1): 9-26
47. Driss T, Vandewalle H, Monod H. Maximal power and force- 64. Cronin JB, McNair PJ, Marshall RN. Velocity specificity, com-
velocity relationship during cycling and cranking exercises in
bination training and sport specific tasks. J Sci Med Sport
volleyball players. J Sports Med Phys Fitness 1998; 38: 286-93
2001; 4 (2): 168-78
48. Kukolj M, Ropret R, Ugarkovic D, et al. Anthropometric,
strength and power predictors of sprinting performance. J 65. Harris GR, Stone MH, O’Bryant HS, et al. Short-term perform-
Sports Med Phys Fitness 1999; 39 (2): 120-2 ance effects of high power, high force, or combined weight-
training methods. J Strength Cond Res 2000; 14 (1): 14-20
49. Bosco C, Rusko H, Hirvonen J. The effect of extra-load condi-
tioning on muscle performance in athletes. Med Sci Exerc 66. McBride JM, Triplett-McBride T, Davie A, et al. The effect of
Sports 1986; 18 (4): 415-9 heavy- vs light-load jump squats on the development of
strength, power, and speed. J Strength Cond Res 2002; 16 (1):
50. Mayhew JL, Piper FC, Schwegler TM, et al. Contributions of
speed, agility and body composition to anaerobic power mea- 75-82
surement in college football players. J Appl Sport Sci Res 67. Moss B, Refsnes PF, Abildgaard A, et al. Effects of maximal
1989; 3 (4): 101-6 effort strength training with different loads on dynamic
51. Meckel Y, Atterbom H, Grodjinovsky A, et al. Physiological strength, cross-sectional area, load-power and load-velocity
characteristics of female 100 metre sprinters of different per- relationships. Eur J Appl Physiol 1997; 75: 193-9
formance levels. J Sports Med Phys Fitness 1995; 35: 169-75
68. Schmidtbleicher D, Haralambie G. Changes in contractile
52. Nesser TW, Latin RW, Berg K, et al. Physiological determi- properties of muscle after strength training in man. Eur J Appl
nants of 40-meter sprint performance in young male athletes. J Physiol Occup Physiol 1981, 228
Strength Cond Res 1996; 10 (4): 263-7
69. Wilson GJ, Newton RU, Murphy AJ, et al. The optimal training
53. Young W, Hawken M, McDonald L. Relationship between load for the development of dynamic athletic performance.
speed, agility and strength qualities in Australian Rules foot- Med Sci Sports Exerc 1993; 25 (11): 1279-86
ball. Strength Cond Coach 1996; 4 (4): 3-6
70. Newton RU, Kraemer WJ, Hakkinen K. Effects of ballistic
54. Schmidtbleicher D. Training for power events. In: Komi PV,
training on preseason preparation of elite volleyball players.
editor. Strength and power in sport. Boston (MA): Blackwell
Scientific Publications, 1992: 381-95 Med Sci Sports Exerc 1999; 31 (2): 323-30

55. Wilson GJ, Lyttle AD, Ostrowski KJ, et al. Assessing dynamic 71. Elliott BC, Wilson GJ, Kerr GK. A biochemical analysis of the
performance: a comparison of rate of force development tests. sticking region in the bench press. Med Sci Sports Exerc 1989;
J Strength Cond Res 1995; 9 (3): 176-81 21 (4): 450-62
56. Young W, McLean B, Ardagna J. Relationship between strength 72. Ritzdorf W. Strength and power training in sport. In: Elliot B,
qualities and sprinting performance. J Sports Med Phys Fitness editor. Training in sport. New York: John Wiley and Sons,
1995; 35 (1): 13-9 1998: 189-237
57. Moritani T. Time course of adaptations during strength and 73. Sale D, MacDougall D. Specificity in strength training: a review
power training. In: Komi PV, editor. Strength and power in for the coach and athlete. Can J Appl Sport Sci 1981; 6 (2):
sport. Boston (MA): Blackwell Scientific Publications, 1992: 87-92
266-78
74. Behm D, Sale D. Intended rather than actual movement velocity
58. Wilson GJ, Murphy AJ. The use of isometric tests of muscular
determines velocity-specific training response. J Appl Physiol
function in athletic assessment. Sports Med 1996; 22 (1):
19-37 1993; 74 (1): 359-68

59. Hakkinen K, Komi PV, Alen M. Effect of explosive type


strength training on isometric force- and relaxation-time, elec-
Correspondence and offprints: Assoc. Prof. John Cronin,
tromyographic and muscle fibre characteristics of leg extensor
muscles. Acta Physiol Scand 1985; 125: 587-600 New Zealand Institute of Sport and Recreation Research,
Auckland University of Technology, Private Bag 92006,
60. Lyttle AD, Wilson GJ, Ostrowski KJ. Enhancing performance:
maximal power versus combined weights and plyometrics Auckland 1020, New Zealand.
training. J Strength Cond Res 1996; 10 (3): 173-9 E-mail: john.cronin@aut.ac.nz

 2005 Adis Data Information BV. All rights reserved. Sports Med 2005; 35 (3)

S-ar putea să vă placă și