Accepted: 18 December 2017

DOI: 10.1111/sms.13045

ORIGINAL ARTICLE

The impact of strength level on adaptations to combined

weightlifting, plyometric, and ballistic training

L. P. James1,2 | G. Gregory Haff3 | V. G. Kelly1,4 | M. J. Connick1 |

B. W. Hoffman5 | E. M. Beckman1

1
School of Human Movement and Nutrition
Sciences, The University of Queensland,
The purpose of this investigation was to determine whether the magnitude of adapta-
Brisbane, QLD, Australia tion to integrated ballistic training is influenced by initial strength level. Such infor-
2
Discipline of Human Movement and Sports mation is needed to inform resistance training guidelines for both higher-­and
Science, Federation University, Ballarat, lower-­level athlete populations. To this end, two groups of distinctly different strength
VIC, Australia
3
levels (stronger: one-­repetition-­maximum (1RM) squat = 2.01 ± 0.15 kg·BM−1;
Centre for Sport and Exercise Science
Research, Edith Cowan University, Perth, weaker: 1.20 ± 0.20 kg·BM−1) completed 10 weeks of resistance training incorporat-
WA, Australia ing weightlifting derivatives, plyometric actions, and ballistic exercises. Testing oc-
4
Brisbane Broncos Rugby League Football curred at pre-­, mid-­, and post-­training. Measures included variables derived from the
Club, Brisbane, QLD, Australia
5
incremental-­load jump squat and the 1RM squat, alongside muscle activity (electro-
School of Health and Wellbeing, University
of Southern Queensland, Toowomba, QLD,
myography), and jump mechanics (force-­time comparisons throughout the entire
Australia movement). The primary outcome variable was peak velocity derived from the un-
loaded jump squat. It was revealed that the stronger group displayed a greater
Correspondence
Lachlan Peter James, School of Human (P = .05) change in peak velocity at mid-­test (baseline: 2.65 ± 0.10 m/s, mid-­test:
Movement and Nutrition Sciences, The 2.80 ± 0.17 m/s) but not post-­test (2.85 ± 0.18 m/s) when compared to the weaker
University of Queensland, Brisbane, QLD,
participants (baseline 2.43 ± 0.09, mid-­test. 2.47 ± 0.11, post-­test: 2.61 ± 0.10 m/s).
Australia.
Email: [email protected] Different changes occurred between groups in the force-­ velocity relationship
(P = .001-­.04) and jump mechanics (P ≤ .05), while only the stronger group dis-
played increases in muscle activation (P = .05). In conclusion, the magnitude of im-
provement in peak velocity was significantly influenced by pre-­existing strength
level in the early stage of training. Changes in the mechanisms underpinning perfor-
mance were less distinct.

KEYWORDS
athletic performance, electromyography, jump squat, neuromuscular, power, resistance training

1 | IN TRO D U C T ION great interest to sports scientists and physical preparation

coaches. These tasks are characterized by an acceleration that
It is understood that maximal impulse-­related expressions continues throughout the entire range of motion (ie weight-
(quantified by measures such as velocity, force, power and lifting actions), often resulting in the athlete or object they
impulse itself), supported by maximal strength, are often are accelerating entering free space (ie plyometric activities
the most important muscular functions driving athletic per- and jump squats). It is theorized that a training plan incorpo-
formance1 and are characteristic of higher-­level competitors rating all of these ballistic modalities under a variety of loads
across a number of sports.2-5 Ballistic exercises are com- is the ideal strategy, as it allows for enhancement through-
monly used to develop these qualities6 and are therefore of out the force-­velocity spectrum and a superior transfer of

1494 | © 2018 John Wiley & Sons A/S. wileyonlinelibrary.com/journal/sms Scand J Med Sci Sports. 2018;28:1494–1505.
Published by John Wiley & Sons Ltd
JAMES et al.   
| 1495

training.7 Furthermore, greater variation in these factors (ie rapid adaptations to such training, underpinned by alterations
loading conditions, modalities, and movement patterns) is in muscle activation, movement mechanics and the force-­
considered advantageous as training experience and strength velocity relationship. Identifying changes that result from
level increases.8,9 Yet, despite its common use by coaches at this stimulus will have a major impact on training practices
the elite level,10-12 little is known about the adaptations to in sport. Furthermore, if differing responses are revealed
such a program design in stronger and weaker individuals. between the two groups, then training interventions can be
It is hypothesized that stronger individuals possess neu- better tailored according to the physiological composition or
romuscular adaptations that form the foundation for an en- development status of the individual.
hanced response to ballistic resistance training.1 Of note,
stronger individuals would generally possess greater neu-
ral drive,13 myofibrillar cross-­sectional area14 and superior
2 | M ATERIAL S AND M ETHO D S
intermuscular coordination.9 As a result, such individuals
are in the later stages of the sequence of events that lead to
2.1 | Experimental design
enhanced maximal ballistic expressions.15,16 Indeed, cross-­ Two groups of distinctly different lower-­body strength levels
sectional investigations have consistently found higher max- (relative 1RM squat) undertook the same 10-­week integrated
imal impulse capabilities in those who are stronger.5,17-21 ballistic training plan for the lower-­body. This prescription
However, the notion of superior adaptive ability among these was divided into two 5-­week training blocks with loading con-
individuals is in contradiction to the principle of diminished ditions and exercise selection based on the principles of perio-
returns22 caused by a history of resistance training, common- dization. Participants attended a single day testing battery at
place in those who are strong. The influence of strength level three separate occasions during the study (baseline, mid, and
on the ability to enhance high-­velocity capabilities is of par- post). Familiarization for all testing and training techniques
ticular relevance to a number of sports as maximal strength occurred across three 1-­hour sessions before baseline testing.
varies between athletes of differing competition levels.5,23,24 Measures derived from the incremental-­load jump squat, in ad-
If an interaction between these factors is present, then differ- dition to strength level and muscle activation, were obtained.
ing training strategies would be required for developmental
and lower-­level athletes compared to those at a higher level.
However, there is limited research into this notion, with the
2.2 | Participants
few experimental investigations comparing the adaptations of Individuals who were male, uninjured, and could competently
stronger and weaker individuals to a ballistic power training perform a back squat were recruited from the university and
intervention failing to produce definitive results.22,25,26 It is surrounding community, resulting in 24 recreationally ac-
possible that the limited variation of the training stimulus (ie tive males who undertook baseline testing. Subjects were
ballistic exercise selection and loading conditions) in these then ranked in accordance with their relative 1RM squat
studies reduced the potential to optimize adaptations in the performance. To establish two groups of distinctly different
stronger participants. strength levels, the 8 middle-­ranked participants were elimi-
Taken together, despite sound theoretical underpinnings, nated. This resulted in a stronger (n = 8; BM = 76.82 ± 6.27;
the experimental evidence available has failed to show signifi- height = 1.72 ± 0.48 m; 1RM squat = 2.01 ± 0.15 kg·BM−1,
cantly greater adaptations to ballistic training in stronger ver- resistance training experience = 4.0 ± 1.31 year) and
sus weaker individuals. Furthermore, because of the limited weaker (n = 8; BM = 82.03 ± 14.7; height = 1.83 ± 0.68 m;
variation in these programs it is not known whether the ad- 1.20 ± 0.2 kg·BM−1; resistance training experience = 1.38
aptations to an integrated approach, such as those commonly 0 ± 0.92 year) strata, thereby enabling between group com-
found in high-­performance settings, would be influenced by parisons. Such a methodology has been previously used to
initial strength level. As the basis for these theorized prefer- form stronger and weaker groups for similar purposes.25
ential adaptations to high-­velocity training among stronger Participant characteristics over the duration of the study are
individuals is a more favorable neuromuscular profile,1 it is presented in Table 1. Written, informed consent was secured
of particular interest to also investigate the changes in neural from all participants, and the study was approved by the
activation and movement mechanics responsible for the per- Bellberry Human Research Ethics Committee, Australia.
formance adaptations. It is therefore the purpose this study
to compare performance changes alongside force-­velocity,
neural, and force-­time responses to an integrated ballistic
2.3 | Training program
training plan between stronger and weaker individuals. Peak Before undertaking training, participants completed three
velocity was chosen as the primary outcome variable due to 1-­hour instructional sessions delivered by the primary inves-
its highly influential contribution to athletic performance.27 tigator, who is certified with both the Australian Strength
It is hypothesized stronger individuals will display more and Conditioning Association and the National Strength and
1496
|    JAMES et al.

TABLE 1 Participant characteristics

IsoSquat/BM Peak velocity Average velocity Jump height

− − −
BM (kg) (N·kg 1) (m·s 1) (m·s 1) (m)
Stronger group
Baseline 76.82 ± 6.27 38.37 ± 6.77 2.65 ± 0.10 1.34 ± 0.10 0.33 ± 0.04
,Ω
Mid-­test 77.49 ± 6.06 41.94 ± 6.20* 2.80 ± 0.17** 1.51 ± 0.14** 0.38 ± 0.05**
Post-­test 77.55 ± 5.94 41.14 ± 5.06 2.85 ± 0.18** 1.52 ± 0.11** 0.37 ± 0.04**
Weaker group
Baseline 82.04 ± 14.07 34.63 ± 5.13 2.43 ± 0.09 1.20 ± 0.12 0.26 ± 0.01
Mid-­test 82.26 ± 14.37 36.56 ± 6.68 2.47 ± 0.11 1.26 ± 0.08 0.28 ± 0.02**
Post-­test 82.24 ± 14.88 39.23 ± 5.41* 2.61 ± 0.10** 1.35 ± 0.06** 0.30 ± 0.02**
BM, Body mass. Velocity and jump height measures are derived from the unloaded jump squat.
*Significant change from baseline (P ≤ .05).
**Significant change from baseline (P ≤ .01).
Ω
Significantly different magnitude of change from the weak group (≤ 0.05).

Conditioning Association. These sessions included detailed Australia). Three minutes of recovery was prescribed be-
coaching on all training and testing activities until profi- tween each set. Subjects performed a general dynamic warm-
ciency was achieved. The training plan included three super- ­up at the beginning of each session (consisting of a series of
vised 1 hour sessions each week over two 5 week blocks squat, lunge and submaximal ballistic actions) and a series of
separated by 1 week to allow for mid-­testing. Workouts were warm-­up sets at progressively increasing loads before each
at least 24 hours apart and consisted of weightlifting deriva- exercise. No additional lower-­body training was permitted
tives, ballistic tasks, and plyometric exercises using a vari- for the duration of the study.
ety of loads. Training emphasis shifted across blocks toward
increased loads for weightlifting derivatives and a decrease
in loading for jump squat actions, in addition to the incor-
2.4 | Testing overview
poration of complex plyometric exercises. Specifically, the Testing sessions were undertaken at week 0 (pre), after week
first block training involved five sets of five repetitions of 5 (mid-­training), and after week 10 (post-­training). Post-­
the power clean and jump squat on day 1 and 3. The power training testing occurred no earlier than 7-­days after the last
clean was performed with 70% 1RM, while the jump squat training session in week 10, and no later than 10 days after,
was performed with 40 and 50% squat 1RM on day 1 and on the basis of the fitness-­fatigue model.25 A week without
3, respectively. On day 2, the hang power clean (55% of the training was allocated following the first block of training
power clean 1RM) and snatch grip pull (70% of the power allowing for mid-­testing to be conducted 3-­5 days after the
clean 1RM) were undertaken across four sets of five repeti- final session of week 5. Each assessment session commenced
tions. During day 1 and 3 of the second block, the loading of with the 1RM squat. The jump squat with an additional
the jump squat was reduced to 0% (day 1) and 30% (day 2), load representing 0% of the 1RM (no added weight), +20%,
while the power clean was increased to 85% 1RM for four +40%, +60%, and +80% of the 1RM squat was undertaken
repetitions across five sets. Additionally, subjects performed in a non-­randomized order to determine maximal neuromus-
the depth jump from a 0.30-­m box using the following sets cular related variables. Finally, the isometric squat at a knee
and repetition scheme: Week 6 – 3 × 3, Week 7 – 3 × 4, angle of 140° was then administered. Simultaneous kinetic
Week 8 – 4 × 4, Week 9 and 10 – 5 × 4. Day 2 of the second (force platform) and electromyography (EMG) readings
block saw an increase in load for both the hang power clean were gathered during the session for selected tests.
and snatch grip pull to 70% and 85% 1RM (of the power
clean), respectively, across five sets of four repetitions. In ad-
dition to this, day 2 included a plyometric rebound split squat
2.5 | Data acquisition procedures
for four sets of three repetitions on each side. Weightlifting
derivatives were encouraged to be performed with maximal
2.5.1 | 1RM squat
intent, while ballistic and plyometric actions were executed A general, followed by a specific dynamic warm-­up was un-
with the goal of maximizing height. Furthermore, during the dertaken before the administration of the 1RM squat. Trials
unloaded jump squat, participants were provided with imme- were then performed until a 1RM was established, with
diate visual and audible peak velocity feedback for each jump each attempt separated by 5-­minutes of passive recovery.25
(GymAware, Kinetic Performance Technology, Canberra, A squat depth to an internal knee angle of <85° of flexion
JAMES et al.   
| 1497

as assessed by 2-­dimensional motion analysis was consid- force platform. Subjects were positioned with a knee angle
ered a successful attempt (stronger: baseline = 82.13 ± 2.42°, of 140° and were instructed to apply maximal force “as hard
mid-­test = 81.31 ± 2.90°, post-­test = 80.5 ± 2.33°; weaker: and as fast as possible” into the immovable bar for 3-­seconds.
baseline = 79.63 ± 5.15°, mid-­test = 81.00 ± 3.63°, post-­ The maximal force value attained during the effort was con-
test = 80.75 ± 2.49°). Data were captured by a Logitech HD sidered force at zero velocity. Strong verbal encouragement
Webcam (model C270, recording at 30fps) positioned 1.5 m was delivered throughout.
to the right of the performer and 0.40 m above the ground.
Processing occurred via Kinovea, (V0.8.15, www.kinovea.
com).
2.5.4 | Muscle activation
Simultaneous surface EMG of the vastus lateralis (VL), VM
(vastus medialis) and biceps femoris (BF) on the right leg was
2.5.2 | Jump squat
acquired during the jump squat, in addition to the isometric
Trials were conducted at a series of ascending relative load- squat. Before placing the EMG electrodes on the site, a razor
ing conditions (+0% of 1RM, ie no additional weight, +20%, was used to shave any hair from the skin. Following this, the site
+40%, +60%, and +80% of the individuals’ 1RM). Participants was then lightly abraded and cleaned to ensure the best quality
were instructed to perform a minimum of two non-­continuous signal from the underlying muscles of interest. A bipolar elec-
jump squats for maximal height using a countermovement to trode configuration was used whereby two stick-­on electrodes
a depth resulting in an internal knee angle of 85° (Kinovea, were placed on the skin slightly distal to the middle of belly
V0.8.15). The jump containing the highest peak velocity in for each of the three muscles. To ensure consistent placement
each loading condition was used for analysis. Three minutes across testing sessions, the location of the electrodes and other
of passive recovery was allowed between each set. landmarks on the leg were traced on to a closefitting elastic gar-
All jump squats were performed on a force platform ment. To further aid in this process, multiple images of the loca-
(Bertec Corporation, Columbus, OH, USA) with the data tions were taken. A 16-­channel wireless EMG system (MYON
sampled at 2000 Hz via a data acquisition device (NI USB-­ 320, Myon AG, Schwarzenberg, Switzerland) was used, and
6259 BNC, National Instruments) and processed using a cus- these data were sampled and saved with the same equipment
tom LabVIEW program (V.12.0f3, National Instruments). and parameters as for the force data above. Data were processed
Data were then saved off-line for secondary processing. using a 6th-­order Butterworth bandpass filter of 50-­300 Hz.
Vertical ground reaction force (Fz) provided direct measures The EMG signal over a 1-­s period of continued maximal force
of force applied to the system. A forward dynamics approach production following the initial peak during the isometric squat
was used via the impulse-­momentum relationship to assess was isolated to perform a root mean square (RMS) with a 50-­
velocity of the center of gravity, while the product of force ms window. This generated the maximal voluntary contraction
and velocity at each time point represented power. Peak (MVC) for all measured muscles. To establish EMG activity
velocity, force, were defined as the greatest instantaneous during the jump squat, the RMS of the EMG signal from the
sample of the respective variable during the action before initiation of the countermovement until takeoff was calculated
takeoff. The velocity and force that occurred at peak power then divided by the time to take off. This was then expressed
in each condition were also established to enable construc- relative to the MVC. Rate of EMG rise (RoR) was calculated
tion of force-­velocity curves. The integral of force with re- as the rate of increase from the minimum to maximum RMS
spect to time for the values exceeding system weight during EMG. These EMG procedures are analogous to those used in
the jump represented impulse. Average power and velocity earlier investigations of muscle activity during jumping.25,28
were calculated from the bottom of the countermovement To directly compare force-­ time curves throughout the
(zero velocity) to takeoff, while rate-­dependent measures jump, individual trials were resampled to an equal number
of force and power were calculated between the respective of frames. This was achieved by adjusting the time delta
minimum and maximum values throughout the movement. between each sample to achieve 300 samples from the ini-
Force, impulse, power, rate of force development, and rate tiation of the countermovement until the participant left the
of power development were divided by system mass to be force platform (The Mathworks, Inc., Natick, MA, USA).
expressed in relative terms. Variables were calculated during Consequently, these 300 samples represented 0%-­100% of
secondary processing via a custom designed Matlab pro- normalized time allowing for point-­by-­point comparison of
gram (The Mathworks, Inc., Natick, MA, USA). force characteristics throughout the action.25,28

2.5.3 | Isometric squat 2.6 | Statistical analysis

After a 10-­minute passive recovery, participants undertook Following confirmation of normality, a repeated measures
an isometric squat in a modified power rack secured over the general linear model was used in conjunction with a post-­hoc
1498
|    JAMES et al.

Bonferroni adjustment to locate any differences between

groups. An alpha level of P ≤ .05 denoted statistical signifi-
cance. A power analyses revealed that for a statistical power
of 80% to be attained, a minimum of 8 participants per group
was needed. To establish practically relevant differences be-
tween means, Cohen’s d effect size (ES) calculations were
employed with thresholds set at <0.2, 0.21-­0.5, 0.51-­0.8 and
>0.8 for trivial, small, moderate, and large magnitudes of
effect, respectively. Data are presented as mean ± standard
deviation. Statistical Package for Social Sciences (Version
23.0, IBM Corporation, Somers, New York, USA) was uti-
lized to analyze non-­magnitude based data, while ES were
calculated using a custom designed spreadsheet (Microsoft
Excel 2013, Microsoft Corporation, Washington, USA).

3 | R ES U LTS

3.1 | Transfer to athletic performance

(0% jump squat, 1RM squat)
All participants across both groups completed 100% of the
required training and testing sessions. The stronger group
possessed a significantly greater 1RM squat than the weaker
group across baseline (stronger: 2.01 ± 0.15 kg·BM−1,
weaker: 1.20 ± 0.2 kg·BM−1, P < .001), mid-­(stronger:
2.06 ± 0.20, weaker: 1.36 ± 0.16 kg·BM−1, P < .001) and
post-­test (stronger: 2.04 ± 0.23, weaker: 1.43 ± 0.15 kg·BM−1,
P < .001). An improvement in this measure was attained by
weaker participants at mid-­and post-­test (P < .001; ES at mid-­ FIGURE 1 Change in peak velocity, average velocity, and
test = 0.84; ES at post-­test = 1.10), while the stronger group’s jump height in the 0% 1RM jump squat between groups at mid-­test
performance remained unchanged. This resulted in a signifi- (A) and post-­test (B). ΩSignificantly different magnitude of change
cantly different change between groups at mid-­(P = .03) and from the weaker group (≤0.05). **Significant change from baseline
post-­test (P = .01). Both groups improved across a number of at ≤0.01
jump variables (Figures 1, 2, Table 2). Of note, the stronger
participants displayed a significantly greater change in peak Conversely, the weaker participants displayed a gradual
velocity at mid-­test than the weaker group. Any significant increase in the contributions of both force and velocity
changes at post-­test across velocity and power variables were across all time points throughout the loading spectrum.
already present at mid-­test among the stronger group. In con- These factors resulted in a greater magnitude of increase
trast, the weak subjects produced changes in these variables at in velocity at peak power in the stronger participants at
post-­test only. A significant decrease in force at peak power mid-­test. In contrast to this, at multiple points throughout
was displayed at post-­test in the stronger participants, while the curve the weak group produced a significantly greater
both groups revealed a significant increase in impulse at mid-­ magnitude of change in force, particularly at the final test-
and post-­test (Table 2). ing point. Accordingly, a similar pattern was found in the
force-­power interaction (Figure 3). In particular, a more
pronounced elevation of this curve (to increased peak
3.2 | Impact on the force–velocity
power values) can be seen in the stronger group at mid-­test
relationship
when compared to the weak participants. This resulted in a
Training resulted in changes to the force-­ velocity and significantly greater magnitude of change in peak power in
force-­power (Figure 3) relationship from baseline in both the high-­force portion of this curve (80% 1RM jump squat
groups. Following a clear rightward and upward shift (to condition) at the mid-­test point when compared to the weak
increased values) of the force-­velocity relationship at mid-­ group. However, at post-­testing the weaker subjects dis-
test in the stronger group, there was a notable regression played a continued shift of this relationship to increased
to lower values under higher-­force conditions at post-­test. peak power values, while a general depression occurred
JAMES et al.   
| 1499

stronger group had significant changes from 3.5% to 27.5%

and 41.5 to 68% of normalized jump duration. At this time
point, the weaker group achieved significant changes from
14.5% to 29.5%, 42% to 61%, and 83% to 90.5% of normal-
ized jump duration. Figure 4 presents these results in graphi-
cal form.

4 | DISCUSSION

This study revealed that a ballistic training plan incorporat-

ing a variety of modalities, movement patterns, and loading
conditions elicited a significantly different performance and
mechanistic response between stronger and weaker partici-
pants over a 10-­week training period. This is of great rel-
evance as such training plans are commonplace in sporting
settings,10,11 and both strength level and high-­velocity capa-
bilities are characteristic of superior athletes across a number
of sports.5,6,23

4.1 | Adaptations in athletic performance

When the extent of improvements was compared between
the two groups, those who were stronger displayed preferen-
tial adaptations to the training stimulus. Alongside a signifi-
cantly superior increase in peak velocity after only 5 weeks,
F I G U R E 2 Change in peak power, average power, and peak the stronger participants displayed significant improvements
force in the 0% 1RM jump squat between groups at mid-­test (A) and at this time point across all velocity-­based variables (peak
post-­test (B). *Significant change from baseline at ≤0.05. velocity, average velocity, and jump height). Following this,
**Significant change from baseline at ≤0.01 the magnitude of improvement across many performance
measures was not markedly different from mid-­test values
among the stronger participants between the mid-­and final after 10 weeks. In contrast, the weaker group did not achieve
testing point. significant improvements in these measures until 10 weeks.
To the authors’ knowledge, this present investigation is the
first to report a statistically greater improvement in a primary
3.3 | Electromyography
outcome measure in stronger versus weaker individuals fol-
No significant differences were present in the change between lowing ballistic training. In large part, this can be attributed
groups for normalized average RMS EMG or RoR across any to the design of the training intervention, whereby a spectrum
of the measured muscles. However, the stronger group dis- of loading conditions, weightlifting derivatives, and plyo-
played a significant increase from pre-­to post-­training in VL metric actions were included, resulting in a potent stimulus
rate of EMG rise. No other significant changes occurred in for adaptation.
either group after training; however, a number of practical The reduced positive adaptations between 5 and 10 weeks
changes existed in both groups (Table 3). Effect size changes in the stronger group were likely a consequence of an inhibi-
in average RMS EMG and RoR were greater across VL and tion in the development of force-­producing capabilities. This
BF in the stronger group at mid-­and post-­test. is indicated by a significant loss of force at peak power after
training, resulting in a significantly different change from the
weaker group. Furthermore, stronger participants displayed
3.4 | Jump mechanics
only trivial changes in peak dynamic force upon completion
At mid-­ test, stronger participants displayed significant of the study (ES: −0.17), while the stimulus produced a large
changes from 17.5% to 25% and 48.0% to 72.5% of normal- effect (ES: 1.04) in this variable among the weaker group.
ized jump duration. Significant alterations among the weaker As stronger individuals generally undertake regular heavy
group at this time point were revealed from 19.5% to 25% and strength training, the cessation of additional lower-­ body
72.5 to 78% of normalized jump duration. After training, the resistance training for the duration of the study was likely
1500
|    JAMES et al.

TABLE 2 Magnitude of change from baseline for performance TABLE 2 (Continued)

variables derived from the 0% 1RM jump squat condition
Stronger group Weak group
Stronger group Weak group
Change from Change from
Change from Change from
baseline (ES, baseline (ES,
baseline (ES, baseline (ES,
95%CI) 95%CI)
95%CI) 95%CI)
Change at post-­test
Velocity
Peak (W·kg−1) 2.69 (0.54, 0.15 to 4.06 (1.13, 0.59
Change at mid-­test
0.92)* to 1.67)**
Peak (m·s−1) 0.15 (0.99, 0.64 to 0.03 (0.35, 0.11
Average (W·kg−1) 2.98 (0.90, 0.48 to 3.37 (1.31, 0.74
1.35)**,Ω to 0.59)
1.32)** to 1.88)**
Average (m·s−1) 0.17 (1.14, 0.85 to 0.06 (0.60, 0.18
RPD (W·kg−1·s−1) 36.18 (0.69, 0.20 50.97 (1.29,
1.43)** to 1.02)
to 1.18) 0.51 to 2.07)**
At peak power (m·s−1) 0.17 (1.13, 0.75 to 0.02 (0.28, 0.27
CI, confidence interval; ES, Cohen’s d effect size.
1.52)**,Ω to 1.47)
*Significant change from baseline at ≤0.05.
Change at post-­test **Significant change from baseline at ≤0.01.
Ω
Peak (m·s−1) 0.21 (1.18, 0.72 to 0.17 (1.35, 0.64 Significantly different magnitude of change from the weak group (≤ 0.05).
ΩΩ
1.63)** to 2.05)** Significantly different magnitude of change from the weak group (≤ 0.01).

Average (m·s−1) 0.18 (1.28, 0.78 to 0.15 (1.25, 0.60

1.79)** to 1.89)** responsible for this response. This highlights the importance
At peak power (m·s ) −1
0.23 (1.30, 0.81 to 0.15 (1.28, 0.52 of maintaining heavy strength training throughout a training
1.79)** to 2.04)** plan, even in those who are already strong and seek to im-
Force prove high-­velocity expressions. Similar findings have been
Change at mid-­test reported when stronger individuals were exposed to jump
Peak (N·kg−1) 0.79 (0.35, −0.63 0.81 (0.79, 0.24 squat training with 0% and 30% 1RM loads.25 However, the
to 1.33) to 1.33) findings of this present investigation also suggest that despite
At peak power −0.44 (−0.35, 0.24 (0.31, 0.07 the inclusion of high-­force actions throughout the interven-
(N·kg−1) −0.98 to 0.28) to 0.36) tion (ie power cleans and snatch grip pulls at 70%-­85% of the
Net impulse 0.49 (1.20, 0.85 to 0.34 (1.15, 0.35 power clean 1RM), force losses during ballistic-­only training
(N·kg−1·s−1) 1.56)** to 1.70)** still occur in stronger individuals. While training had little
RFD (N·kg−1·s−1) 28.94 (1.07, 0.45 12.62 (1.00, impact on the magnitude of dynamic force production in the
to 1.68)** 0.31 to 1.70) strong group, these data suggest that it was the temporal as-
Change at post-­test pects of performance that were most notably influenced. This
Peak (N·kg−1) −0.24 (−0.17, 1.55 (1.04, 0.08 can be seen by the significantly greater improvement in the
−0.83 to 0.48) to 2.00) velocity at peak power than the weaker participants. Because
At peak power −0.88 (−0.68, 0.52 (0.57, 0.05 of the limited change in dynamic force and controlled depth
(N·kg−1) −1.19 to to 1.08) (internal knee angle of 85°), such improvements are likely a
−0.17)*,ΩΩ result of a reduced movement time. In addition, significant
Net impulse 0.77 (1.58, 1.25 to 0.65 (1.51, 1.14 increases in the rate at which force was produced (RFD) were
(N·kg−1·s−1) 1.91)** to 1.89)** present in the stronger group at mid-­test, while the weaker
RFD (N·kg−1·s−1) 25.76 (1.20, 0.68 29.99 (1.34, participants displayed no significant changes in this variable.
to 1.72)** 0.65 to 2.04)** Such temporal-­based factors driving performance enhance-
Power ment in the stronger group is in alignment with previously
Change at mid-­test reported mathematical modeling of power development.15,16
Peak (W·kg−1) 2.5 (0.55, 0.06 to 1.05 (0.31, 0.10 Another notable finding is the greater ES increases in av-
1.05) to 0.53) erage velocity and average power with respect to their peak
Average (W·kg−1) 4.06 (1.01, 0.62 to 1.67 (0.77, 0.24 (instantaneous) variants, which is in contrast to what is pre-
1.40)** to 1.31) viously reported in the literature for single modality training
RPD (W·kg−1·s−1) 62.62 (1.07, 0.45 25.88 (0.87, interventions.25,28 This is particularly pronounced for power
to 1.68)** 0.27 to 1.48) whereby the magnitude of increase in average power (stron-
ger: ES = 1.01; weaker: ES = 0.77) was approximately twice
(Continues) that of peak power (stronger: ES = 0.55; weaker ES = 0.31)
JAMES et al.   
| 1501

F I G U R E 3 (A) Change in the force-­

velocity relationship between groups and
across time points. Measurement points
represent the jump squat with 0%, 20%,
40%, 60%, and 80% of the one-­repetition-­
maximum back squat, alongside peak force
from the isometric squat. ΩΩ Significantly
different magnitude of change from the
alternate group (≤0.01). ΩSignificantly
different magnitude of change from the
alternate group (≤0.05). *Significant change
from baseline at ≤ 0.05. **Significant
change from baseline at ≤0.01. (B)
Change in the force-­power relationship
between groups and across time points.
Ω
Significantly different magnitude of
change from the alternate group (≤0.05).
*Significant change from baseline at
≤ 0.05. Only significant changes with
respect to peak power are indicated.
Significant changes in force at peak power
are presented in Figure 3A

at mid-­test. As instantaneous variables in this study are rep- often occur at different body positions dependent on the lift
resentative of a 0.0005 second epoch, it can be argued that performed, this present training intervention provided an ef-
the average values provide a better indication of the charac- fective stimulus for improved performance throughout the
teristics of the entire movement. Although an instantaneous action, rather than a single point.
velocity will determine the precise spatiotemporal outcome
of a technique, in the case of power it may be more advan-
tageous to produce higher average levels than peak. This is
4.2 | Mechanistic adaptations
because the work done occurs over a period of time, rather
than an instant. For example, decisive actions in sport occur
4.2.1 | Force-­velocity relationship
across epochs of 100-­250 ms29,30 because they require force Training resulted in contrasting shifts of the force-­velocity
to be expressed over some distance for a brief period of time. and force-­power relationships between groups. This is most
What may be responsible for the differences in peak and av- notable in the significantly greater improvements to force
erage values is the variety of movement patterns the subjects at peak power under multiple higher-­force conditions by
were exposed to in this study. As peak velocities and powers the weaker participants. As a result, there was a progressive
1502
|    JAMES et al.

TABLE 3 Magnitude of change from baseline for electromyography (EMG) measures derived from the 0% 1RM jump squat condition

Stronger group Weak group

Average RMS EMG ES (95%CI) Classification ES (95%CI) Classification

Change at mid-­test
VM 0.31 (−0.28 to 0.89) Small 0.17 (−0.42 to 0.76) Trivial
VL 0.53 (−0.25 to 1.31) Moderate 0.42 (−0.32 to 1.15) Small
BF 0.38 (−0.08 to 0.84) Small 0.36 (−0.25 to 0.96) Small
Change at post-­test
VM 0.42 (−0.10 to 0.95) Small −0.12 (−0.79 to 0.54) Trivial
VL 0.72 (−0.04 to 1.40) Moderate 0.65 (−0.33 to 1.62) Moderate
BF 0.68 (−0.01 to 1.36) Moderate 0.54 (−0.05 to 1.13) Moderate
Rate of EMG rise
Change at mid-­test
VM 0.07 (−0.58 to 0.72) Trivial 0.25 (−0.33 to 0.82) Small
VL 0.59 (−0.23 to 1.41) Moderate 0.41 (−0.40 to 1.22) Small
BF 0.42 (−0.16 to 1.0) Small 0.23 (−0.38 to 0.84) Small
Change at post-­test
VM 0.16 (−0.40 to 0.73) Trivial 0.08 (−0.49 to 0.65) Trivial
VL 0.81 (0.15 to 1.46)* Large 0.57 (−0.44 to 1.59) Moderate
BF 0.64 (0.07 to 1.21) Moderate 0.46 (−0.06 to 0.98) Trivial
Magnitude is expressed as Cohen’s d effect sizes (ES) and respective classification.
BF, biceps femoris; CI, confidence interval; RMS, root mean square; VL, vastus lateralis; VM, vastus medialis.

rightward translation of the curve (to increased force) over magnitude of change at post-­ test across all measured
the duration of the study. In contrast to this, the stronger muscles for both RoR and AvRMS. Although the proce-
group exhibited significantly increased velocity capabilities dures in this present investigation cannot determine the
at both extremes of the force-­velocity relationship at mid-­ contribution of motor unit firing frequency or recruitment
test. However, there was a general regression to lower values to this increased muscle activity, previous research has
throughout the curve between 5 and 10 weeks of training, reported extremely large increases in motor unit firing
indicating that a decay of strength occurred. Accordingly, frequency following ballistic training.31 However, further
a significantly greater increase in peak power produced research is required to determine how strength level influ-
under loaded conditions was attained by the stronger group ences these contributions to muscle activation following
at 5 weeks, while this newly attained value was reduced at such a training stimulus. Increases in muscle activation
post-­test. These findings are similar to previous reports of during a sports-­specific movement have been reported
somewhat different force-­velocity responses between high alongside improvements in expressions of power and
and low strength individuals exposed to a ballistic training velocity.25,28,32-35 Low-­
load (thus high-­ velocity) jump
intervention.25 Taken together, this suggests that the ability training has resulted in increases in RoR in both strong
of an individual to operate in different force-­velocity envi- and weak individuals previously.25 The lack of clear re-
ronments following training is influenced by initial strength sponse in this measure among the weaker subjects in this
capabilities. present study may be a reflection of the combined high-­
force/high-­velocity stimulus of the intervention causing
more general adaptations in this group. As training-­
4.2.2 | Muscle activation
induced neural responses differ between a high-­force and
The aforementioned improvements in expressional tim- high-­velocity stimulus,28 it can be expected that when
ing can at least be partly explained by the significant these stimuli are combined there would be a more broad
enhancement of intra-­ muscular activation rates (VL change in these neural measures than would occur in an
RoR) among the stronger group. This is in contrast to isolated loading condition. Although the stronger group
the weaker participants who displayed no significant were exposed to the same training, this represented a
changes. Furthermore, when effect sizes are examined, more velocity-­ dominant stimulus to these participants
stronger individuals also displayed a practically greater (as indicated by the aforementioned changes to their
JAMES et al.   
| 1503

lower-­level strength-­power athletes.37 While there were

little further changes to the second half of the jump (ac-
tive lengthening and concentric phases) at post-­test in this
group, there was a continued improvement in unweighting
(ie a greater reduction in force upon initiation of the coun-
termovement, prior to active lengthening) at the beginning
of the jump. Weaker participants achieved improvements
in similar phases of the jump; however, the epochs were
considerably smaller. This indicates that while both groups
improved their ability to utilize the stretch shortening cycle,
this was achieved to a greater extent by the stronger group in
the early stages of training. However, these improvements
continued through post-­test in the weaker participants re-
sulting considerably greater forces that were achieved ear-
lier in the jump compared to mid-­and pre-­test.

5 | PERSPECTIVE

The present results reveal that the development of maximal

velocity-­based expressions is influenced by pre-­existing
strength level, with greater early stage improvements expe-
rienced by those who are stronger. The mechanisms driv-
ing the changes in performance are also different between
groups, with neural, force-­velocity, and mechanical adap-
tations characteristic of improvements in maximal veloc-
ity in those with already high levels of strength. This is in
contrast to the weaker group who displayed more general
adaptations to training (shifts of both force and velocity
alongside moderate changes in the magnitude and rate of
F I G U R E 4 Changes in the normalized force-­time curve for
the 0%1RM jump squat in the stronger (A) and weaker (B) groups.
muscle activation). These findings are of great value to
*Significant change from baseline to mid-­test at ≤0.05. δSignificant training practices as they reveal that it is advantageous
change from baseline to post-­test at ≤0.05 for individuals to attain a high level of strength before
­emphasizing plyometric actions, weightlifting derivatives,
force-­velocity relationship). This is consistent with find- and ballistic training. However, those who are stronger
ings of increased jump squat RoR in response to low-­load should not remove heavy strength training, as the decay of
ballistic training, while only changes in maximal muscle force-­producing capabilities likely limited improvements,
activation in an isometric squat were experienced in re- particularly after 5 weeks. The results of this study might
sponse to heavy strength training.28 also have important implications for technical training. As
many technical factors are dictated by velocity also,27,38 the
findings of this investigation provide evidence that those
4.2.3 | Jump mechanics who are stronger may more quickly respond to sports spe-
In order for strength to translate into improved jump per- cific training.
formance, the control of force must be optimized.36 The
normalized jump force-­time curves provide valuable infor-
ACKNOWLEDGEMENTS
mation on the characteristics of force application throughout
the movement and therefore explain how jump performance LPJ was supported through an Australian Postgraduate
was achieved. At mid-­test, stronger individuals displayed Award (APA) Scholarship. The authors would like to thank
an increased unweighting (drop into the countermovement, Associate Professor Glen Lichtwark for his input.
before active lengthening). This led to greater eccentric
forces and attainment of peak force earlier in the jump, with
ORCID
no change in its magnitude. These changes are similar to
the distinguishing jump characteristics of higher-­versus L. P. James http://orcid.org/0000-0002-0598-5502
1504
|    JAMES et al.

R E F E R E NC E S 17. Cormie P, McBride JM, McCaulley GO. Power-­time, force-­time,

and velocity-­time curve analysis of the countermovement jump:
1. Cormie P, McGuigan M, Newton R. Developing maximal neu-
impact of training. J Strength Cond Res. 2009;23:177‐186.
romuscular power: part 1 -­Biological basis of maximal power
18. McBride JM, Triplett-McBride T, Davie A, Newton RU. A com-
production. Sports Med. 2011;41:17‐38.
parison of strength and power characteristics between power
2. Young WB, Newton RU, Doyle T, et al. Physiological and anthro-
lifters, Olympic lifters, and sprinters. J Strength Cond Res.
pometric characteristics of starters and non-­starters and playing
1999;13:58‐66.
positions in elite Australian Rules football: a case study. J Sci Med
19. Stone MH, O Bryant HS, McCoy L, Coglianese R, Lehmkuhl M,
Sport. 2005;8:333‐345.
Schilling B. Power and maximum strength relationships during
3. Baker D, Newton R. Comparison of lower body strength, power,
performance of dynamic and static weighted jumps. J Strength
acceleration, speed, agility, and sprint momentum to describe and
Cond Res. 2003;17:140‐147.
compare playing rank among professional rugby league players. J
20. Ugrinowitsch C, Tricoli V, Rodacki A, Batista M, Ricard MD.
Strength Cond Res. 2008;22:153‐158.
Influence of training background on jumping height. J Strength
4. Hansen KT, Cronin JB, Pickering SL, Douglas L. Do force-­
Cond Res. 2007;21:848.
time and power-­time measures in a loaded jump squat differ-
21. Garcia-Pallares J, Lopez-Gullon JM, Muriel X, Diaz A, Izquierdo
entiate between speed performance and playing level in elite
M. Physical fitness factors to predict male Olympic wrestling perfor-
and elite junior rugby union players? J Strength Cond Res.
mance. Eur J Appl Physiol. 2011;111:1747‐1758. Epub 2011/01/12.
2011;25:2382‐2391.
22. Wilson G, Murphy A, Walshe A. Performance benefits from
5. James LP, Beckman EM, Kelly VG, Haff GG. The neuromuscular
weight and plyometric training: effects of initial strength level.
qualities of higher and lower-­level mixed martial arts competitors.
Coaching Sport Sci J. 1997;2:3‐8.
Int J Sports Physiol Perform. 2017;12:612‐620.
23. Suchomel TJ, Nimphius S, Stone MH. The Importance of Muscular
6. Cormie P, McGuigan M, Newton R. Developing maximal neu-
Strength in Athletic Performance. Sports Med. 2016;46:1419‐1499.
romuscular power: Part 2 -­Training considerations for im-
24. James L, Haff G, Kelly V, Beckman E. Towards a determination
proving maximal power production. Sports Med. 2011;41:
of the physiological characteristics distinguishing successful
125‐146.
mixed martial arts athletes: a systematic review of combat sport
7. Newton RU, Kraemer WJ. Developing explosive muscular power:
literature. Sports Med. 2016;46:1525‐1551.
Implications for a mixed methods training strategy. Strength Cond
25. Cormie P, McGuigan MR, Newton RU. Influence of strength on
J. 1994;16:20‐31.
magnitude and mechanisms of adaptation to power training. Med
8. Häkkinen K, Komi PV, Alén M, Kauhanen H. EMG, muscle
Sci Sports Exerc. 2010;42:1566‐1581.
fibre and force production characteristics during a 1 year training
26. Secomb J, Farley O, Lundgren L, Tran T, Nimphius S, Sheppard JM.
period in elite weight-­lifters. Eur J Appl Physiol Occup Physiol.
The effect of initial lower-body strength on the training-specific ad-
1987;56:419‐427.
aptations resulting from combined plyometric and gymnastics train-
9. Sale D. Neural adaptations to strength training. In: Komi PV, ed.
ing. Proceedings of International Conference on Applied Strength
Strength and Power in Sport. Oxford (UK): Blackwell Science;
and Conditioning (ASCA). 2014;2014, Melbourne, Vic, Australia.
2003:281‐313.
27. Winter EM, Abt G, Brookes FC, et al. Misuse of “power” and
10. Ebben WP, Carroll RM, Simenz CJ. Strength and conditioning
other mechanical terms in sport and exercise science research. J
practices of National Hockey League strength and conditioning
Strength Cond Res. 2016;30:292‐300.
coaches. J Strength Cond Res. 2004;18:889.
28. Cormie P, McGuigan MR, Newton RU. Adaptations in athletic
11. Ebben WP, Hintz MJ, Simenz CJ. Strength and conditioning
performance after ballistic power versus strength training. Med
practices of Major League Baseball strength and conditioning
Sci Sports Exerc. 2010;42:1582.
coaches. J Strength Cond Res. 2005;19:538.
29. Tidow G. Aspects of strength training in athletics. New Stud
12. Simenz CJ, Dugan CA, Ebben WP. Strength and conditioning
Athletics. 1990;1:93‐110.
practices of National Basketball Association strength and condi-
30. Andersen LL, Aagaard P. Influence of maximal muscle strength
tioning coaches. J Strength Cond Res. 2005;19:495.
and intrinsic muscle contractile properties on contractile rate of
13. Hakkinen K, Aien M, Komi PV. Changes in isometric force-­and
force development. Eur J Appl Physiol. 2006;96:46‐52.
relaxation-­time, electromyographic and muscle fibre characteris-
31. Cutsem M, Duchateau J, Hainaut K. Changes in single motor unit
tics of human skeletal muscle during strength training and de-
behaviour contribute to the increase in contraction speed after dy-
training. Acta Physiol Scand. 1985;125:573‐585.
namic training in humans. J Physiol. 1998;513:295‐305.
14. Häkkinen K, Komi P, Tesch P. Effect of combined concentric and
32. Jakobsen MD, Sundstrup E, Randers MB, et al. The effect of
eccentric strength training and detraining on force-­time, muscle
strength training, recreational soccer and running exercise on
fiber and metabolic characteristics of leg extensor muscles. Scand
stretch–shortening cycle muscle performance during counter-
J Med Sci Sports. 1981;3:50‐58.
movement jumping. Hum Mov Sci. 2012;31:970‐986.
15. Zamparo P, Minetti A, di Prampero P. Interplay among the
33. Taipale R, Mikkola J, Vesterinen V, Nummela A, Häkkinen
changes of muscle strength, cross-­ sectional area and maxi-
K. Neuromuscular adaptations during combined strength and
mal explosive power: theory and facts. Eur J Appl Physiol.
endurance training in endurance runners: maximal versus ex-
2002;88:193‐202.
plosive strength training or a mix of both. Eur J Appl Physiol.
16. Minetti AE. On the mechanical power of joint extensions as af-
2013;113:325‐335.
fected by the change in muscle force (or cross-­sectional area),
34. Hakkinen K, Komi P. Effect of explosive type strength train-
ceteris paribus. Eur J Appl Physiol. 2002;86:363‐369.
ing on electromyographic and force production characteristics
JAMES et al.   
| 1505

of leg extensor muscles during concentric and various stretch-­ 38. James LP, Robertson S, Haff GG, Beckman EM, Kelly VG. Identifying
shortening cycle exercises. Scand J Sports Sci. 1985;7:65‐76. the performance characteristics of a winning outcome in elite mixed
35. McBride J, Triplett-McBride T, Davie A, Newton R. The ef- martial arts competition. J Sci Med Sport. 2017;20:296‐301.
fect of heavy-­vs. light-­load jump squats on the development
of strength, power, and speed. J Strength Cond Res. 2002;
16:75‐82. How to cite this article: James LP, Gregory Haff G,
36. Bobbert MF, Van Soest AJ. Effects of muscle strengthening on Kelly VG, Connick M, Hoffman B, Beckman EM. The
vertical jump height: a simulation study. Med Sci Sports Exerc. impact of strength level on adaptations to combined
1994;26:1012‐1020.
weightlifting, plyometric, and ballistic training. Scand J
37. McMahon JJ, Murphy S, Rej SJ, Comfort P. Countermovement
jump phase characteristics of senior and academy rugby league
Med Sci Sports. 2018;28:1494–1505. https://doi.
players. Int J Sports Physiol Perform. 2016;12:1‐23. org/10.1111/sms.13045