Leeds Metropolitan University Repository
Paradisis, G.P., Bissas, A., Cooke, C.B. (2009) Combined Uphill and Downhill
Sprint Running Training Is More Efficacious Than Horizontal. International
Journal of Sports Physiology and Performance, 4 (2) June, pp.229-243.
© Human Kinetics 2009. Uploaded in line with the publisher’s self-archiving
policy – see http://www.sherpa.ac.uk/romeo/ for more information.
The Leeds Metropolitan University Repository is a digital collection of
the research output of the University, comprising citation records and,
where copyright permits, full-text articles made available on an Open
For more information, including our policy and submission procedure,
visit the website at http://repository.leedsmet.ac.uk/
Alternatively, email the repository team at firstname.lastname@example.org
International Journal of Sports Physiology and Performance, 2009, 4, 229-243
© 2009 Human Kinetics, Inc.
Combined Uphill and Downhill Sprint
Running Training Is More Efficacious
Giorgos P. Paradisis, Athanassios Bissas,
and Carlton B. Cooke
Purpose: This study examined the effects of sprint running training on sloping sur-
faces (3°) on selected kinematic and physiological variables. Methods: Fifty-four
sport and physical education students were randomly allocated to one of two training
groups (combined uphill–downhill [U+D] and horizontal (H)) and a control group
(C). Pre- and posttraining tests were performed to examine the effects of 8 wk of
training on the maximum running speed (MRS), step rate, step length, step time,
contact time, eccentric and concentric phase of contact time (EP, CP), ight time,
selected posture characteristics of the step cycle, and 6-s maximal cycle sprint test.
Results: MRS, step rate, contact time, and step time were improved signicantly in a
35-m sprint test for the U+D group (P < .01) after training by 4.3%, 4.3%, −5.1%, and
−3.9% respectively, whereas the H group showed smaller improvements, (1.7% (P <
.05), 1.2% (P < .01), 1.7% (P < .01), and 1.2% (P < .01) respectively). There were no
signicant changes in the C group. The posture characteristics and the peak anaerobic
power (AWT) performance did not change with training in any of the groups. Conclu-
sion: The U+D training method was signicantly more effective in improving MRS
and the kinematic characteristics of sprint running than a traditional horizontal train-
Keywords: Kinematic sprinting characteristics, posture characteristics, 6-s
maximal cycle sprint test
Many training methods have been used to improve maximal sprint running
performance by effecting changes in step length and step rate. Running on sloping
surfaces is widely used in training for sprint running.1 Previous studies have
examined kinematic changes of sprinting on a 3° slope and reported an 8.4%
faster maximum running speed (MRS) for the downhill and 2.9% slower MRS for
the uphill slope when compared with horizontal sprinting.2,3 Kunz and Kaufmann4
examined sprinting on a 1.7° slope and reported similar results, whereas Slawinski
Paradisis is with Physical Education and Sport Science, University of Athens, Athens, Greece, and
Bissas and Cooke are with Carnegie Faculty of Sport and Education, Leeds Metropolitan University,
Leeds, United Kingdom.
230 Paradisis, Bissas, and Cooke
et al24 found decreases in MRS, step rate, and step length by 15.6%, 7.4%, and
14.2% respectively when sprinting in 4.9° uphill slope. Positive claims have been
made for the effects of downhill and uphill training on the kinematic characteris-
tics on horizontal running,1,5 but only two studies have reported experimental data.
Tziortzis6 showed that after 12 wk of training on a downhill slope of 8° the MRS
increased by 2.1% and the step length increased by 1.4%, whereas the step rate did
not change. Paradisis and Cooke7 have reported that after 6 wk of training on a
downhill slope of 3° the MRS increased by 1.1% and the step rate increased by
2.3%, whereas the step length did not change. For uphill training, Tziortzis6
reported that the MRS and step rate increased by 3.3% and 2.4% respectively,
although the changes in step length were not statistically signicant, whereas Par-
adisis and Cooke7 reported no statistically signicant changes after the uphill
There have also been positive claims for the benets of training on combined
uphill and downhill sloping surfaces, although again these claims have not been
substantiated with published experimental data.1,5 Only Paradisis and Cooke7
have assessed the effects of 6 wk combined uphill–downhill sprinting training, on
sloping surfaces of 3° and showed improvements on MRS and step rate by 3.5%
and 3.4% respectively. In addition, the horizontal training and control groups did
not produce any statistically signicant changes.
The aim of this study was to evaluate further the effects of 8 wk of training on
combined uphill and downhill sloping surfaces of 3° compared with both training
on the horizontal and a control group in terms of the kinematic and posture char-
acteristics of sprinting and performance in the 6-s maximal cycle sprint test
(MCST). The current study will, therefore, either conrm or refute the ndings of
the previous preliminary study7 using more appropriate group sizes and provide a
comparison of training effects for 8 wk with those reported for 6 wk.
Fifty-four male sport and physical education students participated in this study
(age 24.1 ± 2.1 years, mass 75.3 ± 10.2 kg, height 1.75 ± 0.08 m). All subjects
were active in different sports but none was a sprinter; their mean MRS was 8.20
± 0.74 m·s−1. However, to participate in this study, all subjects were asked to ter-
minate any other sport activity. Informed consent was obtained from each partici-
pant before data collection, where the study was granted with ethics approval by
the appropriate board of the university. A wooden uphill–downhill platform was
used and it was covered with synthetic track surface. The width of the platform
was 1.20 m and the total distance covered was 80 m: 20 m horizontal, 20 m uphill
at 3° slope, 10 m horizontal, 20 m downhill at 3° slope, and 10 m horizontal
The participants were randomly assigned to three groups:
• U+Dwastrainedontheuphill-downhillplatform(n = 18)
• Hwastrainedonthehorizontal(n = 18)
Figure 1 — The uphill–downhill platform.
232 Paradisis, Bissas, and Cooke
• Cwasthecontrolgroupanddidnottrain(n = 18)
After completion of a 20-min warm-up, both training groups performed 6
80 m sprints at maximal intensity per session, three times a week, where the time
between repetitions (10 min) was sufcient for the participants to recover fully.8
This training program continued until the fourth week, after which one repetition
was added for both training groups, for each of the remaining 4 wk (training ses-
sions for the last week were 10 80 m). Group C maintained their normal physi-
cal activities throughout the experimental period without performing any kind of
Pre- and posttraining tests were employed to evaluate the effects of training on the
kinematic and posture characteristics of sprinting and AWT performance. The
sprints were performed in a corridor 60 m long and 2.5 m wide in the biomechan-
ics laboratory, and the oor was covered with a synthetic track surface (tartan) 55
m long and 2.5 m wide. The corridor was well lit and the ambient temperature was
25o C. After completion of a 20-min warm-up, the participants performed three
maximal sprint runs over a 35-m distance using a standing start. The time between
the repetitions (10 min) was sufcient for the participants to recover fully.8 The
adoption of three trials for each participant was to establish the magnitude of vari-
ability associated with repeated trials.
A Kodak EktaPro 1000 high-speed video camera was used to collect record-
ings of the sagittal plane of a full stride (two consecutive steps) of all three maxi-
mal sprint runs, sampling at 250 Hz. Filming was performed with the camera
placed at the 35-m distance (so it should be near to MRS as evidence from the
literature has showed that MRS is achieved at about 30 m9,10) and 10 m from the
performance plane such that its optical axis was approximately horizontal, form-
ing an angle of 90° with the horizontal plane of running. For the digitization pro-
cess, a metal calibration frame (2 2 m) was lmed such that the x-axis was
parallel to the horizontal and the y-axis was perpendicular to the horizontal.
Analysis of the Video Data
The hardware of the digitizing system comprised a video projector Imager LCD
15E (General Electronic, USA), a TDS Graphic tablet and controller (x,y resolu-
tion, 0.025 mm; active area 1.20 0.90 m), interfaced with an IBM computer that
ran the digitizing program DIGIT (Leeds Metropolitan University). A standard
17-point,11 14-segment model of the human performer based on the data of Demp-
ster12 was used to represent the human performer and to calculate the position of
the center of mass. Reliability of the digitizing process was established in previ-
ous study3 by repeated digitizing of one sprinting sequence at the same sampling
frequency with an intervening period of 48 h. Contact time, ight time, step time,
step length, ight distance, step rate, and MRS were calculated according to meth-
ods reported previously.3 The comparison of left and right foot contact times was
performed using the limits of agreement method (calculating the mean ± s of the
differences between left and right feet, where the boundaries of agreement based
on the expression ± 1.96).13 Additionally, the following were calculated
Combined Uphill and Downhill Sprint Training 233
according to methods reported previously3: the touchdown and take-off angles of
the knee (), hip (), shank to running surface (), trunk to running surface (;
trunk angle was determined by the line between the hip and glenohumeral joints
of the right side of the body), and the distance parallel to the running surface
between a line perpendicular to the running surface that passes through the center
of mass and the contact point at touchdown (DCM TD) and at take-off (DCM TO;
6-s Maximal Cycle Sprint Test
A 6-s maximal cycle sprint test (MCST) was used to determine the peak anaerobic
power and consisted of a 6-s maximal sprint on a modied cycle ergometer
(Monark 814E) against a braking force of 0.075 kg·kg−1 of body mass. This test
was included to establish whether any adaptations to training transferred to a dif-
ferent mode of exercise than that used in sprint running training. These data will
be useful in characterizing the specic and general responses to sprint training
using running on sloping surfaces. Initially, the participants were instructed to
perform a warm-up activity for 5 min by cycling at 60 rpm with 1.5 kg of load.
After a 5-min rest period, each participant performed three all-out trials and the
best of the three trials was analyzed. The participants were instructed to attain an
Figure 2 — Location of the body landmarks and visualization of the angles: knee (), hip
(), shank to running surface (), trunk to running surface (), and the distance parallel to
the running surface between a line perpendicular to the running surface that passes through
the center of mass and the contact point at touchdown and takeoff. Note that in terms of
simplicity the ipsilateral and contralateral hip/glenohumeral joints appeared to be in the
same position; however, this was not the case.
234 Paradisis, Bissas, and Cooke
initial pedaling frequency of 80 rpm with 0.5 kg of resistance. When this pedal
rate was achieved, the load was applied and the participants accelerated, pedaling
maximally for 6 s. The time between repetitions (10 min) was sufcient for the
participants to recover fully.8
A three-way ANOVA with repeated measures on two factors (trial and test) was
used to establish if there were any signicant differences between the trials, the
tests (pre and post) and the groups (training groups) and any interaction effects.
Each dependent variable was analyzed using a separate ANOVA. A multivariate
analysis of variance, used to analyze all dependent variables, was not completed
as there were insufcient participants for the required degrees of freedom. In the
event of signicant main effects, a post hoc Tukey test was used to locate the dif-
ferences. The signicance level for the tests was set at P < .05.
Comparison of the Three Trials
To assess the consistency between the three trials, a comparison was performed
across the groups. Factors that could affect the consistency include fatigue, lack of
familiarization, boredom, natural variation, insufcient warm-up, and lack of
motivation. There was no signicant difference in all the analyzed variables
between the three trials for all the groups.
Comparison of Left and Right Leg
In the analysis of the pre- and posttraining tests, contact time was measured from
the left foot throughout. This was justied through a comparison of left and right
foot contact times using the limits of agreement method.13 The mean ± s of the
differences between left and right feet was 0.001 ± 0.003 s and the boundaries of
agreement were −0.008 and 0.005 s (heteroscedasticity correlation was close to
zero). Given these results it was concluded that there were no signicant differ-
ences between the contact times for the left and right foot.
Effects of Different Training Methods
Kinematic Characteristics. MRS increased signicantly after 8 weeks of train-
ing for the U+D group by 4.3% and for the H group by 1.7%, whereas for the
control group did not change signicantly (Table 1). In the U+D group, all partici-
pants produced increases in the MRS (0.35 ± 0.21 m·s−1, ranged from 0.10 m·s−1
to 0.89 m·s−1), whereas in the H group thirteen participants increased their MRS
(0.21 ± 0.15 m·s−1, ranged from 0.05 m·s−1 to 0.53 m·s−1). Finally, the repeated-
measures ANOVA showed no signicant differences between the groups for all
the pretraining tests.
Similarly, step rate increased signicantly for U+D group (4.3%) and H
group (1.2%), whereas it did not change signicantly for the C group (Table 1).
Table 1 Mean ± SD of the three trials (post- to pretraining values) of the kinematic characteristics of all groups
MRS (m·s−1) SR (Hz) SL (m) CT (ms) FT (ms) ST (ms)
U+D Pre 8.25 ± 0.69 3.98 ± 0.32 2.07 ± 0.11 128 ± 18 125 ± 11 253 ± 20
Post 8.60 ± 0.68 4.15 ± 0.38 2.08 ± 0.15 121 ± 15 121 ± 12 243 ± 22
P0.001 0.001 0.763 0.001 0.052 0.001
CI 0.28 to 0.43 0.11 to 0.24 −0.037 to 0.028 4.3 to 8.6 0.9 to 7.5 6.1 to 13.8
HPre 8.12 ± 0.40 3.91 ± 0.14 2.02 ± 0.08 128 ± 11 128 ± 10 256 ± 10
Post 8.26 ± 0.42 3.96 ± 0.17 2.03 ± 0.07 125 ± 11 127 ± 10 253 ± 11
P0.010 0.001 0.396 0.001 0.175 0.001
CI 0.03 to 0.23 0.02 to 0.07 −0.032 to 0.013 1.0 to 3.3 −0.4 to 2.3 1.5 to 4.6
CPre 8.20 ± 0.86 4.05 ± 0.20 1.99 ± 0.18 125 ± 6 122 ± 9 247 ± 13
Post 8.16 ± 0.81 4.04 ± 0.20 1.98 ± 0.18 126 ± 5 123 ± 9 248 ± 13
P0.180 0.200 0.887 0.508 0.439 0.058
CI −0.019 to 0.096 −0.009 to 0.041 −0.012 to 0.013 −1.4 to 0.7 −2.2 to 1.0 0.2 to 6.0
Abbreviations: U+D = combined uphill and downhill training group, H = horizontal training group, C = control group, CI = condence interval, MRS = maximum
running speed, SR = step rate, SL = step length, CT = contact time, FT = ight time and ST = step.
236 Paradisis, Bissas, and Cooke
Fifteen participants increased their step rate for the U+D (0.23 ± 0.10 Hz, ranged
from 0.04 Hz to 0.39 Hz), whereas in the H group 14 participants increased their
step rate (0.06 ± 0.05 Hz, ranged from 0.01 Hz to 0.15 Hz).
The contact time decreased signicantly for U+D group (5.1%) and H group
(1.7%) after the 8 weeks of training, whereas it did not change signicantly for the
C group (Table 1). Sixteen participants reduced their ight time in the U+D group
(8 ± 5 ms, range = 1 to 19 ms), whereas 15 participants reduced it in the H group
(5 ± 4 ms, range = 1 to 15 ms).
Step time decreased signicantly for U+D group by 3.9% and for H group by
1.2%, whereas for the C group it was not signicantly different (Table 1). Fifteen
participants shortened their step time for the U+D group (14 ± 6 ms, range = 3 to
23 ms), whereas 14 participants shortened it for the H group (4 ± 3 ms, range = 1
to 9 ms).
Finally, step length remained unaltered for U+D, H and C groups (Table 1),
the ight time showed a trend toward a decrease by 3.1% for the U+D group after
the 8 weeks of training but this was not statistically signicant. The step length for
the H and C groups remained unaltered (Table 1).
Concentric and Eccentric Phases of Contact. The concentric phase of the
contact time decreased signicantly for the U+D group after the 8 weeks of train-
ing (11.5%), whereas for the H and C groups it did not change signicantly (Table
2). There were no signicant changes in the eccentric phase of the contact time for
all groups, after the 8 weeks of training (Table 2).
Posture Characteristics. There was generally a small effect on the posture char-
acteristics for touchdown and take-off after the 8 weeks of training. The U+D
group showed signicant changes in the knee (3°) and shank (3°) angles for the
contact phase and the hip angle (6°) for takeoff after the 8 weeks of training,
whereas the H group showed signicant changes in the hip angle during the con-
tact phase by 3° and during the takeoff phase by 2°. The C group did not show any
signicant changes (Table 3 and Table 4).
Table 2 Mean ± SD (post- to pretraining values) of the eccentric
and concentric phases of all groups
U+D H C
EP (ms) Pre 53 ± 9 56 ± 7 57 ± 8
Post 53 ± 9 55 ± 6 56 ± 9
P0.954 0.745 0.456
CI −4.1 to 3.9 −4.4 to 6.0 −1.9 to 4.1
CP (ms) Pre 75 ± 19 70 ± 12 67 ± 10
Post 67 ± 17 69 ± 14 69 ± 10
P0.003 0.614 0.069
CI 3.5 to 14.0 −3.4 to 5.5 −3.6 to 0.1
Abbreviations: U+D = combined uphill and downhill training group, H = horizontal training group, C
= control group, CI = condence interval, EP = eccentric phase of contact time, CP = concentric phase
of contact time.
Table 3 Mean ± SD (post- to pretraining values) of the posture characteristics at contact
Knee (°) Hip (°) Shank (°) Trunk (°)DCM (m)
U+D Pre 144 ± 7.4 135 ± 7.0 91 ± 5.4 82 ± 4.8 0.30 ± 0.06
Post 147 ± 7.2 132 ± 5.7 94 ± 4.7 80 ± 4.9 0.32 ± 0.04
P0.001 0.091 0.001 0.183 0.069
CI 1.56 to 4.27 −0.44 to 5.47 1.24 to 4.85 −1.02 to 4.95 −0.048 to 0.002
HPre 151 ± 6.2 134 ± 5.5 92 ± 3.9 78 ± 3.3 0.30 ± 0.03
Post 149 ± 5.1 137 ± 3.8 92 ± 4.6 78 ± 3.8 0.29 ± 0.02
P0.244 0.0012 0.479 0.663 0.836
CI −1.36 to 4.93 1.08 to 5.78 −2.23 to 1.10 −3.49 to 2.29 −0.06 to 0.07
CPre 146 ± 2.0 133 ± 3.7 92 ± 3.5 77 ± 2.7 0.30 ± 0.03
Post 147 ± 3.1 134 ± 4.0 93 ± 4.0 78 ± 3.7 0.30 ± 0.04
P0.078 0.101 0.055 0.096 0.678
CI −1.77 to 0.11 −1.60 to 0.16 −1.71 to 0.02 −1.53 to 0.14 −0.01 to 0.02
Abbreviations: U+D = combined uphill and downhill training group, H = horizontal training group, C = control group, CI = condence interval, DCM = the distance
parallel to the running surface between a line perpendicular to the running surface that passes through the center of mass and the contact point.
Table 4 Mean ± SD (post- to pretraining values) of the posture characteristics at takeoff
Knee (°) Hip (°) Shank (°) Trunk (°)DCM (m)
U+D Pre 164 ± 6.5 207 ± 7.8 42 ± 4.7 84 ± 5.4 0.60 ± 0.06
Post 162 ± 8.6 201 ± 6.6 42 ± 5.1 82 ± 5.4 0.61 ± 0.06
P0.194 0.001 0.705 0.087 0.486
CI −1.03 to 4.72 1.97 to 8.52 −0.96 to 1.38 −0.46 to 6.23 −0.03 to 0.01
HPre 164 ± 5.5 203 ± 5.1 43 ± 3.2 84 ± 2.0 0.60 ± 0.04
Post 163 ± 7.1 205 ± 2.9 43 ± 3.3 84 ± 2.2 0.60 ± 0.03
P0.162 0.002 0.811 0.832 0.250
CI −0.68 to 3.68 0.27 to 3.48 −1.58 to 1.99 −1.34 to 1.10 −0.01 to 0.02
CPre 164 ± 4.1 204 ± 4.3 42 ± 1.2 83 ± 0.9 0.60 ± 0.05
Post 165 ± 5.3 203 ± 3.4 42 ± 1.2 83 ± 1.6 0.58 ± 0.05
P0.411 0.245 0.053 0.347 0.056
CI −2.91 to 1.26 −0.39 to 1.39 −1.26 to 0.01 −0.35 to 0.94 −0.01 to 0.02
Abbreviations: U+D = combined uphill and downhill training group, H = horizontal training group, C = control group, CI = condence interval, DCM = the distance
parallel to the running surface between a line perpendicular to the running surface that passes through the center of mass and the contact point.
Combined Uphill and Downhill Sprint Training 239
Peak Anaerobic Power. The results of the best trial of the 6-s MCST showed no
signicant differences between the pre- and posttraining tests for the MCST for
any group (from 1207.7 ± 172.9 to 1219.3 ± 187.3 W for U+D, from 1085.4 ±
188.9 to 1098.3 ± 198.8 W for H group and 1067.3 ±1076.7 ± 291.9 W for the C
group). These ndings suggested that the training had not increased the ability to
generate a higher peak anaerobic power output in an alternative mode of
The methodological procedures used for digitization and calculation of kinematic
and posture variables, for the comparison of repeated trials and for the comparison
of the left and right step shown to be consistent enough for the effective compari-
son of adaptations to various sprint training methods against a control group.
There were no signicant differences between the pre- and posttraining tests for
all the analyzed variables in the C group, where other studies6,14,15 reported similar
results. The results of the current study were not inuenced by a learning effect,
which means that the familiarization of the subjects before the pretraining test was
sufcient. Therefore, it can be argued that if any pre- to posttesting changes
occurred, these could be attributed as the effect of the training.
The H training method produced signicant increases in MRS (1.7%) and
step rate (1.2%), whereas contact time decreased (1.7%) as did step time (1.2%)
after training, with only minor changes in posture characteristics. Dintiman16 after
8 wk of horizontal training, observed an improvement of 5.2% in performance for
50 m, whereas Suellentrop17 found a 2.5% improvement in 100-m performance
after 6 wk of training, but there are no experimental data regarding changes in step
rate, step length, contact time, and ight time, in the literature. However, as the
correlation between MRS and performance is very high (r = .90)18,19 it can be
concluded that the current study has produced ndings that are consistent with
those predicted by the limited literature for subjects of similar level of expertise.
The results of this study showed that traditional horizontal training produced
small improvements in step rate, contact and step time, variables that inuence
The U+D training produced an increase in MRS of 4.3%, which were accom-
panied by an increase in step rate by 4.3%, whereas the step length did not change.
The increase in step rate was mainly due to a shorter step time (−3.9%), which
was affected by the shorter contact time (−5.1%). The U+D training produced an
11.5% decrease in the concentric phase of contact time after training, whereas the
eccentric phase did not show any signicant changes. This is arguably the most
important adaptation to training, which may account for the improvement in run-
ning speed. In addition, the shortening of the concentric (propulsive) phase and
effectively the shortening of contact time could be interpreted as an improvement
of muscle power.1,21,25 However, in the context of this study the suggestion of
improvements in the force-time (power) muscle’s characteristics is hypothetical,
as no measurement of power was conducted. In addition, the lack of changes in
the eccentric phase is rather surprising, as a reduction of this phase was expected.
Slawinski et al showed the vastus lateralis was less active but for a longer time
240 Paradisis, Bissas, and Cooke
during the concentric phase in uphill sprinting of ~3° (MRS 6.28 ± 0.38 m·s−1),
whereas no differences occurred during the eccentric phase.24 However, Gottschall
and Kram showed a decrease in eccentric impulse and an increase in the concen-
tric impulse during similar uphill sprinting.29 The role of eccentric and concentric
phases in the improvement of MRS as well as the changes of the muscles activa-
tion during uphill and downhill sprinting needs move evaluation.
Despite the signicant changes that occurred in almost all the kinematic vari-
ables after the training period, U+D training did not produce signicant changes
in the posture characteristics. The only exception was an increase in the shank
angle of 3° at contact, which can be explained by the increase of the knee angle
(3°) and a decrease in the hip angle at take off by 6°, all of which can be explained
by the decrease of the concentric phase. It can therefore be concluded that the
combined uphill–downhill training method did not signicantly alter the subject’s
Overall, the superiority of the U+D training method was clear from the
results, with statistical analysis demonstrating that the improvements produced
were signicantly greater than for the H training method. There are few reports in
the literature concerning the effects of U+D training methods, suggesting that a
combination of training methods (uphill and downhill) should produce better
results than any other training method1 and indicating that a combination of uphill
and downhill training would produce signicant improvements in all the kine-
matic characteristics of sprint running.20–22 These suggestions are supported by
the ndings of the current study, which showed that the combined method of
training on the uphill, horizontal, and downhill produced signicant improve-
ments in almost all the kinematic variable analyzed. In contrary, the results indi-
cated that the training employed did not improve performance in the 6-s maximal
cycle sprint test. It seems that the generation of forces to produce peak power in
the 6-s test is based on a different adaptation to that seen in the sprint running
groups, which was not stimulated with the specic sprint running training regi-
men used in the current study.
It can be argued that a faster sprinting speed could be attained by shortening
the step time while the step length remained the same. The step time could be
shortened by reducing the contact time and keeping the ight time the same, as
was the case for the U+D group. However, if the contact time was shortened and
the muscle force remained the same, the impulse produced by the muscles would
decrease (I = F t). In such a scenario, the gain from a shortened contact time
should be lost by producing a smaller step length, but step length did not change
in the U+D group. So, as the impulse would be the same and the contact time
shortened, the muscle force must be increased to produce a higher step velocity.
As the MRS was increased, the contact time was shortened and the step length
remained the same in the U+D group, it could be hypothesized that muscle force
was improved. This hypothesis is partially supported by Wood’s23 conclusion that
to increase MRS, athletes should increase the muscle force of the hamstring. Stud-
ies have demonstrated enhanced muscular loading applied to the hip, knee and
ankle extensors25–28 during uphill running (in lower range of speed ~4.5 m·s-1),
whereas Slawinski24 showed a decreased activation of the hamstrings muscles
during contact time (running at 6.28 m·s−1) in ~3° uphill running; however, no
data are available regarding muscle activation during downhill running. It seems
Combined Uphill and Downhill Sprint Training 241
that there may be a link between the force-time characteristics of the muscles and
the production of shorter contact time and eventually the production of greater
MRS, but this needs further evaluation, since in the context of this study no mea-
surement of muscle force was conducted.
A comparison of the ndings from the current study with those previously
published7 showed a greater magnitude of training response in the current study,
in similar subject expertise ( the MRS, step rate, contact time, and step time were
improved for the U+D group by 4.3%, 4.3%, 5.1%, and 3.9% respectively, whereas
in the previous study7 the MRS, step rate, and step time were increased by 3.5%,
3.4%, and 3.3% respectively, but the contact time did not change signicantly). It
is possible that the greater magnitude of change in the current study might be
partly due to the longer training period, as the subjects’ level of expertise was
similar (8 wk vs. 6 wk in the previous study7). This suggests that the specic train-
ing adaptations associated with the combined uphill–downhill methods continue
while the training stimulus is applied. However, this particular interpretation must
be made with caution, since the only way such a claim can be objectively evalu-
ated is to monitor the training adaptations longitudinally throughout the training
program. Further work is therefore required to substantiate this suggestion, but the
magnitude of training response for the U+D method is certainly encouraging in
comparison with horizontal sprint training.
During running on the platform, subjects experience a 20-m resistive stimu-
lus (uphill), followed by a 10-m normal stimulus (horizontal) and after that a 20-m
facilitative stimulus (downhill). During the resistive stimulus the neuromuscular
system will be overloaded owing to extra resistance (5% of the body weight
because of the 3° slope).7 By repetitive application for a certain time, the body
will adapt to that extra load and as a result some trends of change in the MRS and
kinematic characteristics of horizontal running occur. However, in downhill, an
extra propulsive force (5% of the body weight because of the 3° slope) produces
a supramaximal speed.7 During the uphill part of the platform, the MRS would be
reduced by 2.9% whereas during the downhill part the MRS would be increased
by 8.4%, producing a net increased in the average running speed, over the whole
distance (80 m), compared with maximum horizontal running.2 With training, the
body adapts to this stimulus and increases MRS by improving some of the kine-
matic characteristics. The results of both the previous7 and present studies suggest
that this quick transition from the rst stimulus to the second, from one form of
overload to another, beneted the neuromuscular system. The immediate transi-
tion from the overload status to the facilitated status seems to be a key factor in
enhancing the training adaptation. However, to investigate some of the possible
mechanisms that produce this adaptation further work is needed. It is important to
identify the effects of training on the maximum force and the force-time charac-
teristics from the dominant muscles during sprinting, to have some information on
possible cause and effect.
242 Paradisis, Bissas, and Cooke
The U+D training method was signicantly more effective in improving the maxi-
mum sprinting speed and the associated kinematic characteristics of sprint run-
ning in active sports subjects than an equivalent horizontal training method, with
little change in running posture. The correlation coefcient between MRS and
resulting performance in the 100 m was reported as 0.90 and 0.96 for male and
female sprinters respectively, indicating the importance of the maximum speed
for high-level performance.30,31 Similarly, Tziortzis6 found a correlation 0.88
between MRS and resulting performance. Susanka et al,31 interpreting these
results, reported that MRS seems to be the most important factor in male sprinters
in the 100-m race. So, it could be speculated that the combined uphill–downhill
training method is more effective in improving performance in short distance
sprinting events. This study therefore provides further objective evidence substan-
tiating the efcacy of the combined U+D training method for improving maxi-
mum horizontal sprinting speed, which is important in a range of sports, including
athletics and a variety of major team games.
1. Dintiman G. Sprinting: What research tells the coach about. Washington DC:
AAHPER publications; 1974.
2. Paradisis G, Cooke CB, Bissas A. Sloping surface sprinting kinematics and running
posture. J Sports Sci. 1988;16:13–14.
3. Paradisis G, Cooke CB. Kinematic and postural characteristics of sprint running of
sloping surfaces. J Sports Sci. 2001;19:149–159.
4. Kunz H, Kaufmann D. Biomechanics of hill sprinting. Track Technique. 1981;82:2603–
5. Costello F. Resisted and assisted training to improve speed. Track Field Q Rev.
6. Tziortzis S. Effects of training methods in sprinting performance. Doctoral Disserta-
tion. University of Athens, Dept. of Physical Education and Sport Science, Athens,
7. Paradisis G, Cooke CB. The effects of sprint running training on sloping surfaces. J
Strength Cond Res. 2006;20(4):767–777.
8. McArdle W, Katch F, Katch V. Exercise Physiology: Energy, Nutrition and Human
Performance. London: Lea & Febiger; 1991.
9. Murase Y, Hoshikawa T, Yasuda N, Ikegami Y, Matsui H. Analysis of the changes in
progressive speed during 100 meter dash. In: Komi PV, ed. Biomechanics V-A. Balti-
more: University Park Press; 1976:200–207.
10. Moravec P, Ruzicka J, Susanka P, Dostal E, Kodejs V, Nosek M. The 1987 Inter-
national Athletic Foundation / IAAF Scientic Project Report: time analysis of the
100 metres events at the 2nd World Championships in Athletics. New Stud Athletics.
11. Plagenhoef S. Patterns of Human Motion: a cinematographic analysis. New Jersey:
Prentice Hall, Inc; 1971.
12. Dempster WT. Space requirements of the seated operator. WADC Technical Report.
Dayton, OH: Wright-Patterson Air Force Base; 1955.
13. Bland JM, Altman G. Statistical methods for assessing agreement between two meth-
ods of clinical measurement. Lancet. 1986;i:307–308.
Combined Uphill and Downhill Sprint Training 243
14. Gutoski FP. Effects of force treadmill sprint training on selected physiological param-
eters. Doctoral Dissertation. University of Edmonton, Alberta; 1974.
15. Irwin D. A study of stride length and stride rate changes after high speed treadmill and
sprint training. Master Dissertation. University of Alberta; 1974.
16. Dintiman G. Effects of various training programs on running speed. Res Q Exerc
17. Suellentrop JM. A variation of Russian Downhill sprint training for selected College
students. Master Dissertation. Southern Illinois University at Carbondale; 1979.
18. Komi PV. Stretch-shortening cycle. In: Komi PV, ed. Strength and Power in Sport.
London: Blackwell; 1992:169–179.
19. Pauletto, B. Speed-power training. Scholastic coach. 1993;63:54-55.
20. Milakov M, Cox V. Improving speed by training on sloping surfaces. Track Technique.
21. Mero A, Komi PV. Effects of supramaximal velocity on biomechanical variables in
sprinting. Int J Sport Biomech. 1985;1:240–252.
22. Verkhoshanky YV. Speed training for high level athletes. New Stud Athletics.
23. Wood GA. Biomechanical limitations to sprint running. Med Sport Sci. 1987;25:58–
24. Slawinski J, Dorel S, Hug F, et al. Elite long sprint running: A comparison between
incline and level training sessions. Med Sci Sports Exerc. 2008;40:1155–1162.
25. Delecluse C, Van Coppenolle H, Willems E, Van Leemputte M, Diels R, Goris M.
Inuence of high-resistance and high-velocity training on sprint performance. Med
Sci Sports Exerc. 1995;27:1203–1209.
26. Roberts TJ, Belliveau RA. Sources of mechanical power for uphill running in humans.
J Exp Biol. 2005;208:1963–1970.
27. Sloniger MA, Cureton KJ, Prior BM, Evans EM. Anaerobic capacity and muscle acti-
vation during horizontal and uphill running. J Appl Physiol. 1997;83:262–269.
28. Sloniger MA, Cureton KJ, Prior BM, Evans EM. Lower extremity muscle activation
during horizontal and uphill running. J Appl Physiol. 1997;83:2073–2079.
29. Gottschall JS, Kram R. Ground reaction forces during downhill and uphill running. J
30. Baumann W, Schwirtz A, Gross V. Biomechanik des Kurzstreckenlaufs. In: Ballreich
A, Kuhlow A, eds. Biomeckanik der Leichtathletik. Stuttgart: Enke Verlag; 1986:1–
31. Susanka P, Moravec P, Dostal E, et al. Fundamental motor abilities and selected bio-
mechanical variables related to performance in 100 m OG Seoul 1988. Monaco: IAF;