reliability, usefulness, and validity of a repeated sprint ... · pdf fileyoung elite women...

3

ORIGINAL INVESTIGATIONSInternational Journal of Sports Physiology and Performance, 2010, 5, 3-17© Human Kinetics, Inc.

Buchheit is with the Laboratory of Exercise Physiology and Rehabilitation, Faculty of Sport Sciences, University of Picardie, Jules Verne, Amiens, France. Spencer is with Jerv Football Club, Grimstad, Norway. Ahmaidi is with the Laboratory of Exercise Physiology and Rehabilitation, Faculty of Sport Sciences, University of Picardie, Jules Verne, Amiens, France.

Reliability, Usefulness, and Validity of a Repeated Sprint and Jump Ability Test

Martin Buchheit, Matt Spencer, and Said Ahmaidi

Purpose: Two studies involving 122 handball players were conducted to assess the reliability, usefulness, and validity of a repeated shuttle-sprint and jump ability (RSSJA) test. The test consisted of 6 × (2 × 12.5-m) sprints departing on 25 s, with a countermovement jump performed during recovery between sprints. Methods: For the reliability and usefulness study, 14 well-trained male handball players performed the RSSJA test 7 d apart. Reliability of the test variables was assessed by the typical error of measurement, expressed as a coefficient of variation (CV). The minimal changes likely to be “real” in sprint time and jump power were also calculated. For the validity study, players of seven teams (national to international levels, women and men) performed the RSSJA test. Results: CV values for best and mean sprint time were 1.0% (90% CL, 0.7 to 1.6) and 1.0% (90% CL, 0.7 to 1.4). CV values for best and mean jump peak power were 1.7% (90% CL, 1.2 to 2.7) and 1.5% (90% CL, 1.1 to 2.5). The percent sprint and jump decrements were less reliable, with CVs of 22.3% (90% CL, 15.7 to 38.3) and 34.8% (90% CL, 24.2 to 61.8). Minimal changes likely to be “real” for mean sprint time and jumping peak power were –2.6% and 4.8%. Qualitative analysis revealed that the majority of between-team differences were rated as “almost certain” (ie, 100% probability that the true differences were meaningful) for mean sprint and jump performances. Conclusion: The RSSJA test is reliable and valid to assess repeated explosive effort sequences in team sports such as handball. Test results are likely to be representative of gender and competition level; thus the test could be used to discriminate across playing standards and monitor fitness levels.

Keywords: handball, speed testing, jumping ability, agility, repeated efforts sequence

In team sports such as basketball, handball, or netball, players have to repeat sequences of short explosive efforts, such as sprints (<15 m) with frequent changes in direction1–4 followed by maximal jumping movements. Furthermore, planned jumps occur predominantly after high-intensity runs, such as throwing in the air after a fast strike in handball.2 In addition, repeated sprint ability has been shown to be associated

4 Buchheit, Spencer, and Ahmaidi

to actual match performance5 and competitive level of play.6 Vertical jump height/power has also been considered a relevant performance index in team sports7 and is thus considered as a discriminating variable of different competitive standard.8

Relatively few studies have evaluated repeated sprint ability in the field,9–12 while assessment of jump ability is generally only investigated as an isolated qual-ity.8,13 Although assessment of repeated sprint and jump abilities have previously been integrated,14 the assessment of repeated effort sequences, specific of team sports, has been poorly investigated and has been restricted to volleyball-specific movements.4 Thus, to integrate the explosive effort sequences to simultaneously assess repeated sprint and jump abilities, we proposed15 to add a universal coun-termovement jump during the recovery phase (within 3 s following the preceding sprint) of a commonly used shuttle-based repeated sprint ability test.6,10 While examining the effect of different training regimens on performance of this test in young elite women handball players, we observed relative changes in repeated jump ability to be significantly higher than those in single jumping, single sprinting, and repeated sprint abilities.15 This suggested that repeated jump ability, especially when jumps are performed immediately after maximal sprints, such as in real games, might be a highly sensitive measure to track changes in athletic performance fol-lowing a training period.15

Despite these encouraging results,15 investigating the reliability and validity of the test was still warranted. Moreover, extending the practicability of the test to confidently monitor the progression of a player, which is generally defined as the test “usefulness,” was also required to increase the relevance the test. Hopkins16 has pro-posed to compare the magnitude of the smallest worthwhile change in performance with the uncertainty or noise in the test result. Knowledge of the minimal change in performance likely to be “real”17 has also been shown to be of great importance in a training perspective. While over-ground repeated sprint ability tests have been shown to discriminate soccer players performance across different ages and levels of competition,6,18 it is not known whether the repeated sprint and jump test is also likely to differentiate handball players of various age, gender, and competitive play of a similar extent. Normative data are not yet available for this test, which might be of great interest for the development of talent selection and recruitment purposes.

The aim of this study was thus to (1) investigate the reliability of this repeated shuttle-sprint and jump ability test, (2) determine its usefulness and practicality in the field to monitor the progression of an athlete, and (3) examine its construct validity, that is, whether repeated sprinting and jumping performances are repre-sentative of age, gender, and competitive play level.

Methods

Subjects and Study Design

One hundred twenty-two players from four women’s and three men’s French handball teams participated in the study (Table 1). Their maturational status was estimated via self-assessment questionnaires administrated by an experienced investigator.19 Training details, such as weekly training hours, were also collected. All subjects played at a competitive level, from national leagues to international levels. Fourteen players from a team competing in the 5th French male league

Repeated Sprint and Jump Ability Test 5

participated in both the reliability and construct validity study, whereas all other athletes participated in the second study only. After one familiarization trial the preceding week,20 all tests were performed at the end of the preparatory phase, 1 wk before the start of the competitive season. All tests were preceded by a standardized warm-up including athletic drills and accelerations, which was terminated with a single sprint-jump sequence performed at maximal intensity. Testing began 2 min 30 s after this last effort, which was used as the players’ reference performance. The players involved in the reliability study performed the test twice, separated by 7 d. The sample size used in the reliability study was consistent with that used in previous reliability studies in the field.11,20 All players were provided with the procedures and risks associated with participation in the study and gave their written informed consent before participation. The study was approved by the university’s human research ethics committee.

Repeated Shuttle-Sprint and Jump Ability Test (RSSJA). The RSSJA was designed to simulate the repeated explosive efforts commonly performed in team sports.1–3 To compare athletic profiles of players of different team sports on a single protocol, we chose a test that used common sprinting distances and 180° changes of direction, as well as a universal countermovement jump (CMJ). The test consisted of six maximal 2 × 12.5-m shuttle sprints (≈5 s) departing every 25 s. During the ≈20-s recovery between sprints, subjects had to decelerate, perform a CMJ, and then an active recovery (covering 36 m ≈ running at 2.1 m⋅s–1; Figure 1). Since an arm swing may complicate comparisons between pre- and posttraining blocks, due to a change in upper body strength affecting the contribution of arm swing to CMJ performance, participants were ask to keep their hands on their hips during all jumps. The standardization of arm position during jumping is obviously important for test reliability. The depth of the countermovement jumps were self-selected to minimize intervention and thereby maximize the potential application to practical settings where time limitations may exist. Moreover, it is apparent that any change in countermovement depth does not influence jumping height.21 All athletes were verbally encouraged throughout the test and asked to jump as high as possible. Two seconds before starting each sprint, the subjects were asked to assume the ready position (ie, center of gravity up to the front foot, placed 5 cm before the first timing gate) and await the start signal from a compact disc.10 The best sprint time (RS

best;

s) and jump height (RJbest

; cm), the mean sprint time (RSmean

; s), and jump height (RJ

mean; cm) were calculated. Peak power during CMJ was also calculated according

to the following formula, which has been shown to be the most accurate to estimate peak power from jumping height, in both women and men:22 CMJ (W) = (60.7 × h [cm]) + (45.3 × m [kg]) – 2055. The percent sprint time (RS

dec; %) and percent

jump height or peak power (RJdec

; %) decrement were calculated, respectively, as ((mean sprint time / best sprint time × 100) – 100) and (100 – (mean jump / best jump × 100)). Sprinting times were recorded with photoelectric cells (Wireless Timing Radio Controlled, Brower Timing System, Colorado, USA), and jumping height with an Optojump (Microgate, Bolzano, Italy).10

Statistical Analysis. The distribution of each variable was examined with the Shapiro-Wilk normality tests. Homogeneity of variance was verified by a Levene test, and data herein are presented as means and standard deviations (± SD). To examine the reliability of the test over the two consecutive trials, pairwise

6

Tab

le 1

S

ub

ject

ch

arac

teri

stic

s

Gro

upn

Age

(y)

Tann

er s

tage

Sta

ture

(cm

)W

eigh

t (kg

)Tr

aini

ng p

ract

ice

(h ·

wk–1

)

Reg

iona

l elit

e yo

ung

wom

en’s

te

am17

15.9

± 1

.1II

I =

1, I

V =

13,

V =

316

8.1

± 5

.363

.8 ±

6.8

≈10

Wom

en’s

Fre

nch

natio

nal u

nder

16

yr

team

2915

.8 ±

0.6

III

= 5

, IV

= 2

2, V

= 2

175.

7 ±

6.3

68.7

± 8

.4≈1

0

Wom

en’s

Fre

nch

natio

nal u

nder

18

yr

team

2017

.5 ±

0.5

III

= 3

, IV

= 1

3, V

= 4

174.

7 ±

6.2

69.2

± 8

.7≈1

0

Wom

en’s

2nd

leag

ue te

am14

26.1

± 5

.1IV

= 2

, V =

12

174.

8 ±

6.6

71.4

± 9

.2≈9

Reg

iona

l elit

e yo

ung

men

’s te

am14

15.6

± 1

.1II

I =

3, I

V =

8, V

= 3

182.

6 ±

4.7

75.8

± 8

.4≈1

0

Men

’s 5

th le

ague

1423

.1 ±

4.2

IV =

1, V

= 1

318

2.2

± 4

.677

± 8

.1≈5

Men

’s 2

nd le

ague

1426

.5 ±

4.7

IV =

1, V

= 1

318

4.4

± 7

.584

.4 ±

10.

8≈1

2

Not

e. V

alue

s ar

e m

ean

±SD

. Age

, pub

erta

l dev

elop

men

t (Ta

nner

sta

ge),

ant

hrop

omet

ric

para

met

ers,

and

trai

ning

pra

ctic

e of

ath

lete

s pl

ayin

g fo

r th

e se

ven

team

s.


comparisons were first applied to determine any learning effect or systematic bias with paired t test. The magnitude of differences between consecutive trials was also expressed as standardized mean difference (Cohen effect sizes, ES). The criteria to interpret the magnitude of the ES were as follows: <0.2 trivial, 0.2 to 0.5 small, 0.5 to 0.8 moderate, >0.8 large.23 The spreadsheet of Hopkins24 was also used to determine the change in the mean between trials and the typical error of measurement (TE, s or cm), expressed as a coefficient of variation (CV, %). Numerous earlier studies have reported biomechanical variables with CVs below 5% as reliable.9,11–13 It is important to acknowledge that having the best reliability does not mean a variable is the most “useful” at measuring something valuable, as a number of physiological measures have high reliability but some are not sensitive measurement tools.23 As a result, a CV <5% was set as the criterion to declare a variable as reliable. The usefulness of the test was assessed while comparing the SWC (0.2 multiplied by the between-subject deviation, based on Cohen’s effect size principle) with the typical error.23 If the typical error was below the SWC, the test was rated as “good”; if the typical error was similar to the SWC, the rating was “OK”; and if the typical error was higher that the SWC, a rating of “marginal” was given.16 The “smallest difference needed to be considered real” (MD, corresponding to a change likely to be “almost certain”) was calculated as TE × 1.96 × √2.17 To examine the construct validity of the test, possible differences in the reliable parameters of the various playing levels were assessed with a one-way ANOVA (“team” factor with seven levels). When a significant main effect was identified, a Tukey post hoc test was used to further delineate differences between teams (Minitab 14.1 Software, Minitab Inc, Paris, France). For all analyses, the level of significance was set at P < 0.05. In addition to this null-hypothesis testing, teams’ performance were also assessed for clinical significance using an approach based on the magnitudes of differences.25 Between-team differences in mean sprint times and jumping peak powers were thus assessed using 90% confidence limits and the chance that the true (unknown) values for a given team were better or poorer than these for the six

Figure 1 — Schematic of the shuttle sprint- and jump-based repeated effort ability test.


other teams was calculated.25 The quantitative chances for a team of having better or poorer performance were assessed qualitatively as follows: <1%, almost certainly not; 1% to 5%, very unlikely; 5% to 25%, unlikely; 25% to 75%, possible; 75% to 95%, likely; 95 to 99, very likely; >99%, almost certain. If the chance of having better and poorer performances were both >5%, the true difference was assessed as unclear. For all pairwise comparisons, thresholds for clinical differences were calculated as 0.2 × the pooled SD of the two groups of interest.

Results

Maximal Effort at the Start of the Repeated Sequences

Best sprint time and jump height recorded during the RSSJA test were 100.1 ± 1.8% and 98.9 ± 2.2% of the reference performances undertaken once before the test, respectively. There was also no significant difference between performances during the test and the reference trial (P = .39 and .26 for best sprint time and jump height, respectively).

Short-Term Reliability

The pairwise analysis revealed no significant differences between the two trials (all P > .05), regardless of the performance variable assessed. Differences in all indices between repeated trials displayed “trivial” ES. All other reliability variables are presented in Table 2. Values for TE and CV were low for best and mean sprint time and jump height/power. Conversely, percent sprint and jump decrements displayed very high CV values (>20%). Coefficients of variation were consistently lower for jumping ability when expressed in watts compared with centimeters.

Test Usefulness

Changes in performance likely to be “real” are presented in Table 2.

Construct Validity

Since %Dec measures displayed poor reliability for both sprint and jump perfor-mance, this variable was not used in the construct validity analysis. Figures 2 and 3 illustrate sprint and jump performance of the test in the seven different teams. The one-way ANOVA revealed a significant team effect (all P < .001), with teams ranked in the “expected” order. Between-team differences highlighted by the post hoc analysis are given in Figures 2 and 3. Magnitude-based inferences for dif-ferences between teams/levels in mean sprint times and peak jumping power are presented in Tables 3 and 4, respectively. The qualitative analysis revealed that all between-team differences were rated as “almost certain” for mean sprint times, with the exception of the Women’s 2nd league team vs the Regional elite young men’s team comparison (Table 3). Furthermore, all between-team differences were rated as “almost certain” for mean peak jumping power, with the exception of the Women’s 2nd league team vs Women’s French national under 18 yr team and Regional elite young men’s team vs Men’s 5th league comparisons, where differences were “unclear” (Table 4).

9

Tab

le 2

M

easu

res

of

relia

bili

ty fo

r re

pea

ted

sp

rin

ts a

nd

jum

ps

du

rin

g t

he

RS

SJA

tes

t in

th

e 5t

h F

ren

ch m

en’s

h

and

bal

l lea

gu

e RS

best

RS

mea

nR

S%

Dec

RJ be

stR

J mea

nR

J %D

ec

Hei

ght

Pow

erH

eigh

tPo

wer

Hei

ght

Pow

er

TE

(90%

CL

)

0.05

s

(0.0

4–0.

08)

0.05

s

(0.0

4–0.

08)

0.61

%

(0.4

4–0.

98)

1 cm

(0.7

–1.6

)

58 W

(42–

94)

0.9

cm

(0.7

–1.5

)

55 W

(40–

89)

2.0%

(1.5

–3.3

)

1.6%

(1.2

–2.6

)

CV

(90%

CL

)

1.0%

(0.7

–1.4

)

1.0%

(0.7

–1.6

)

22.3

%

(15.

7–38

.3)

2.9%

(2.1

–4.7

)

1.7%

(1.2

–2.7

)

2.9%

(2.1

–4.7

)

1.5%

(1.1

–2.5

)

21.2

%

(14.

9–36

.2)

34.8

%

(24.

2–61

.8)

Dif

fere

nce

(90%

CL

)

0.05

s

(0.0

1–0.

08)

0.02

s

(–0.

02–0

.06)

–0.4

8%

(–0.

97–0

.02)

–0.5

cm

(–1.

3–0.

3)

–32

W

(–72

– 16

)

–1.4

cm

(–2.

1 to

–0.

6)

–82

W

(–12

7– 3

7)

–0.2

%

(–1.

9–1.

4)

–0.1

%

(–1.

5 to

–1.

2)

ES

(rat

ing)

0.01

(Tri

vial

)

0.01

(Tri

vial

)

–0.1

2

(Tri

vial

)

–0.1

(Tri

vial

)

–0.0

4

(Tri

vial

)

0.0

(Tri

vial

)

–0.1

(Tri

vial

)

–0.1

(Tri

vial

)

–0.0

3

(Tri

vial

)

SWC

(%)

(tes

t rat

ing)

0.05

s

(1.1

%)

(OK

)

0.05

s

(1.0

%)

(OK

)

0.22

%

(8.9

%)

(mar

gina

l)

0.9

cm

(2.8

%)

(OK

)

131

W

(3.8

%)

(goo

d)

0.9

cm

(3.5

%)

(goo

d)

125

W

(3.9

%)

(goo

d)

0.7%

(6.4

%)

(mar

gina

l)

0.5%

(8.1

%)

(mar

gina

l)

MD

–0.1

3 s

(–2.

63%

)

–0.1

3 s

(–2.

63%

)

–1.6

8%

(–45

.5%

)

+2.

7 cm

(+7.

9%)

+16

1 W

(+4.

6%)

+2.

5 cm

(+8.

5%)

+15

3 W

(+4.

8%)

–5.8

%

(–48

.4%

)

–4.5

%

(–63

.8%

)

Not

e. T

ypic

al e

rror

of

mea

sure

men

t (T

E),

TE

exp

ress

ed a

s a

coef

ficie

nt o

f va

riat

ion

(CV

), d

iffe

renc

e in

mea

n be

twee

n th

e tw

o tr

ial,

effe

ct s

ize

(ES)

and

ES

ratin

g (s

ee

Met

hods

), s

mal

lest

wor

thw

hile

cha

nge

(SW

C)

and

ratin

g of

the

test

(se

e M

etho

ds),

min

imal

dif

fere

nce

need

ed to

be

cons

ider

ed a

s “r

eal”

(M

D)

calc

ulat

ed f

or b

est s

prin

t tim

e (R

S best;

s) a

nd j

ump

heig

ht (

RJ be

st;

cent

imet

ers

and

wat

ts),

mea

n sp

rint

tim

e (R

Sm

ean;

s)

and

jum

p he

ight

(R

J mea

n; c

entim

eter

s an

d w

atts

), a

nd p

erce

nt s

prin

t tim

e (R

S %D

ec; %

) an

d ju

mp

heig

ht (

RS %

Dec

; %)

decr

emen

t.


DiscussionThe main findings of the current study reveal that best and mean sprinting and jump-ing performances during RSSJA showed good reliability, whereas the decrement scores showed poor reliability. Regarding the usefulness of this test to monitor the progression of an athlete, we estimated that performance improvements of at least 2.6% and 4.8% in mean sprint time and estimated jump peak power have to be detected to attain an “almost certain beneficial” increase in performance. Moreover, the test showed construct validity since test results were representative of gender and competitive play level.

Test Reliability

The reliability of the repeated shuttle-sprint and jump ability test was very good with the absence of any systematic bias (ie, the between-test differences were rated as “trivial” based on effect size calculation) and CV values approximately 1.0% for sprints. The CV for mean repeated sprint performance was similar to that reported previously (0.7%,11 0.8%,9 and 2.4%4). This confirms that jumping during recov-ery has no likely adverse effect on repeated sprint reliability.4 The current study

Figure 2 — Best (RSb) and mean (RS

m) sprint times for athletes playing in the seven dif-

ferent teams. (a) Significant difference (P < .05) vs Women’s French national U-16, (b) vs Women’s French national U-18, (c) vs Women’s 2nd league, (d) vs Regional elite young mean, (e) Men’s 5th league, and (f) Men’s 2nd league. Values are mean ± SD.


examined, for the first time, the reliability of repeated countermovement jumping performance, when conducted immediately after maximal sprints with changes of directions as is evident during competition.1–3 A learning effect was not appar-ent for best and mean jumping height, but CV values (3%) tended to be slightly higher than for sprint performance. Conversely, estimated peak power values led to lower CV values (1.5% for mean CMJ peak power), which were similar to those of repeated sprints. This apparent higher reproducibility of peak power compared with jumping height was in fact related to the subject’s body mass, which is used as a constant in peak power calculation22 and is not likely to change considerably between measurements. Thus, the lower CV observed for power values does not account for an accurate lower error of measurement and it is more likely that the ability to detect change relative to the random error is approximately the same in both instances. Since jumping height and peak power are two measures commonly used in the field to assess jumping capacities, we intended to provide CV and normative values for both measurements. Nevertheless, CV values in the current study were slightly higher than those reported for repeated countermovement jumps without prior sprints (1.9%)13 or for a volleyball-specific test in elite Australian players (0.5%).4 We can hypothesize that the subjects’ training background, sporting activity, explosive efforts sequence type, or jump height measurement device might

Figure 3 — Best (RJb) and mean (RJ

m) jumping power (W) for athletes playing in the seven

different teams. (a) Significant difference (P < .05) vs Women’s French national U-16, (b) vs Women’s French national U-18, (c) vs Women’s 2nd league, (d) vs Regional elite young mean, (e) Men’s 5th league, and (f) Men’s 2nd league. Values are mean ± SD.

12

Tab

le 3

M

agn

itud

e-b

ased

infe

ren

ces

for

mea

n d

iffer

ence

s in

mea

n s

pri

nt

tim

e as

a f

un

ctio

n o

f g

end

er, a

ge,

an

d p

layi

ng

leve

l

Reg

iona

l el

ite y

oung

w

omen

’s

team

Wom

en’s

Fr

ench

nat

iona

l un

der

16 y

te

am

Wom

en’s

Fr

ench

nat

iona

l un

der

18 y

te

amW

omen

’s 2

nd

leag

ue te

am

Reg

iona

l elit

e yo

ung

men

’s

team

Men

’s 5

th

leag

ueM

en’s

2nd

le

ague

Reg

iona

l elit

e yo

ung

wom

en’s

team

—–5

.1%

(–7.

3, –

2.7)

0/0/

100

–8.1

%

(–10

.4, –

5.8)

0/0/

100

–12.

2%

(–15

, –9.

3)

0/0/

100

–12.

8%

(–15

.1, –

10.4

)

0/0/

100

–14.

3%

(–16

.3, –

12.2

)

0/0/

100

–17.

8%

(–19

.8, –

15.8

)

0/0/

100

Wom

en’s

Fre

nch

natio

nal u

nder

16

y te

am

——

–3.2

%

(–5.

5, –

0.9)

0/5/

94

–7.5

%

(–10

.3, –

4.6)

0/0/

100

–8.1

%

(–10

.5, –

5.7)

0/0/

100

–9.7

%

(–11

.7, –

7.6)

0/0/

100

–13.

4%

(–15

.3, –

11.4

)

0/0/

100

Wom

en’s

Fre

nch

natio

nal u

nder

18

y te

am

——

—–4

.4%

(–7.

4, –

1.3)

0/4/

96

–5.1

%

(–7.

6, –

2.5)

0/1/

99

–6.7

%

(–8.

8, –

4.5)

0/0/

100

–10.

5%

(–12

.6, –

8.4)

0/0/

100

Wom

en’s

2nd

leag

ue

team

——

——

–0.7

%

(–3.

9, 2

.7)

19/3

9/42

–2.4

%

(–5.

3, 0

.7)

4/18

/78

–6.4

%

(–9.

2, –

3.5)

0/0/

100

Reg

iona

l elit

e yo

ung

men

’s te

am—

——

——

–1.7

%

(–4.

2, 0

.9)

5/23

/72

–5.8

%

(–8.

2, –

3.3)

0/0/

100

Men

’s 5

th le

ague

——

——

——

–4.1

%

(–6.

2, –

2.0)

0/1/

99

Not

e. V

alue

s re

pres

ent m

ean

diff

eren

ce (

90%

CL

) and

per

cent

age

chan

ce o

f hav

ing

bette

r/tr

ivia

l/poo

rer v

alue

s th

an th

e ot

her t

eam

s ob

tain

ed fr

om q

ualit

ativ

e an

alys

is.

13

Tab

le 4

M

agn

itud

e-b

ased

infe

ren

ces

for

mea

n d

iffer

ence

s in

mea

n p

eak

jum

pin

g p

ow

er a

s a

fun

ctio

n o

f g

end

er, a

ge,

an

d p

layi

ng

leve

l

Reg

iona

l el

ite y

oung

w

omen

’s te

am

Wom

en’s

Fre

nch

natio

nal u

nder

16

y te

am

Wom

en’s

Fre

nch

natio

nal u

nder

18

y te

am

Wom

en’s

2n

d le

ague

te

am

Reg

iona

l el

ite y

oung

m

en’s

team

Men

’s 5

th

leag

ueM

en’s

2nd

le

ague

Reg

iona

l elit

e yo

ung

wom

en’s

te

am

—–9

.3%

(–15

.8, –

2.2)

0/6/

93

–14.

3%

(–20

.8, –

7.3)

0/1/

99

–14.

3%

(–21

.3, –

6.7)

0/1/

99

–27.

3%

(–34

.0, –

19.9

)

0/0/

100

–31.

2%

(–36

.8, –

25.0

)

0/0/

100

–43.

9%

(–47

.9, –

39.7

)

0/0/

100

Wom

en’s

Fre

nch

natio

nal u

nder

16

y te

am

——

–5.5

%

(–11

.8, 1

.3)

2/23

/75

–5.5

%

(–12

.6, 2

)

4/23

/74

–19.

8%

(–26

.7, –

12.3

)

0/0/

100

–24.

1%

(–29

.7, –

18.0

)

0/0/

100

–38.

2%

(–41

.9, –

34.2

)

0/0/

100

Wom

en’s

Fre

nch

natio

nal u

nder

18

y te

am

——

—–0

.1%

(–7.

8, 8

.4)

28/4

4/28

–15.

2%

(–22

.7, –

6.9)

0/1/

99

–19.

7%

(–25

.9, –

12.9

)

0/0/

100

–34.

6%

(–38

.8, –

30.1

)

0/0/

100

Wom

en’s

2nd

le

ague

team

——

——

–15.

1%

(–23

.1, –

6.4)

0/1/

99

–19.

7%

(–26

.4, –

12.3

)

0/0/

100

–34.

6%

(–39

.3, –

29.5

)

0/0/

100

Reg

iona

l elit

e yo

ung

men

’s

team

——

——

—–5

.3%

(–14

.2, 4

.4)

7/27

/66

–22.

9%

(–29

.4, –

15.8

)

0/0/

100

Men

’s 5

th le

ague

——

——

——

–18.

6%

(–24

.5, –

12.2

)

0/0/

100

Not

e. V

alue

s re

pres

ent m

ean

diff

eren

ce (9

0% C

L)

and

perc

enta

ge c

hanc

e of

hav

ing

bette

r/tr

ivia

l/poo

rer v

alue

s th

an th

e ot

her t

eam

s ob

tain

ed fr

om q

ualit

ativ

e an

alys

is.


explain these slight differences. Even though the use of a force plate would have been more precise to measure jumping power,14 we believe that the methodology used here to measure jumping height via any field device (eg, Optojump, vertex, yardstick, or contact matt, and valid conversion equations for both genders22) is more accessible and affordable in the field for coaches and trainers. The poor reli-ability of performance decrement in mean sprinting times and jumping power were comparable to sprint data reported previously (14.9%11 and 30.2%9). The current study also shows for the first time that the reliability of repeated jump decrement is poor, regardless of the measure used (ie, height vs power). Finally, the sample size used here to assess the reliability of the test could be considered as small.23 However, the fact that we found good reliability suggests that increasing the sample size would not have had a much greater effect on the results.

Test UsefulnessBecause typical error was not higher than that calculated for the smallest worthwhile change, the usefulness of the test was rated as “OK” and “good” for best and mean sprint time and jump height (Table 2). From a practical viewpoint, to ascertain meaningful or “almost certain” changes in performance, it is therefore suggested that a performance improvement of at least 2.6% (–0.13 s) and 4.8% (+153 W), or 8.5% (+2.5 cm), in mean sprint time and jump peak power (or height) have to be detected to attain a “real”17 increase in performance. Calculation of this “minimal difference needed to be real” can be considered as very conservative compared with other approaches.16 Nevertheless, interpreting changes in performance based on smaller threshold values, as the “smallest worthwhile change,” for example half of a CV or TE × √2, would require the use of specifically designed spreadsheets, which may be impractical in the field setting. We thus propose to provide coaches with thresholds that were “almost certain” (approximately 100%) to be representative of a true change/difference in performance. The calculated minimal values in the current study are consistent with the magnitude of changes in repeated shuttle-sprint performance observed after a short training period in elite junior soccer (–2.2%9), elite adolescents (–2.0%10 and –3.9%26) and adult handball (–2.8%27) players. Conversely, in accordance with Spencer,11 percent decrement of sprint (and jump) were likely to be worthless. The test was rated as “marginal” for this index, because of the considerable noise within this measurement variable.

Test ValidityAs expected, the younger women playing at the lower handball level displayed significantly and “almost certainly” the lowest repeated sprint (Figure 2 and Table 3) and jumping (Figure 3 and Table 4) abilities when compared with the women’s teams of a higher level and all of the men’s teams. Conversely, the adult men playing in the highest league had significantly and “almost certainly” the best performances when compared with all other teams. The fact that magnitude-based inferences showed meaningful differences in performance between almost all teams (Table 3 and 4), but not the ANOVA and post hoc analyses (Figure 2 and 3), confirms that a null-hypothesis-based approach might be too conservative and is not likely to be the best method to assess differences in performance in sport science.25,28 Even though the current study is the first to report measures of overall repeated


explosive sequences in team sport players of different age, gender, and competi-tive play levels, the present data are in accordance with previous observations that have shown strength, (repeated) speed, power, and aerobic qualities to differ between gender and competitive levels in handball,29 volleyball,4 basketball,30 and soccer.6,8 For instance, Mujika et al8 reported age- and gender-specific differences in performance on several explosiveness-related field tests such as vertical jump, 15-m sprint time, or agility test, with younger women tending to display the worst performance. Furthermore, professional Italian soccer players have been shown to have better repeated sprint performance than amateur players.6 Similar results were also reported by Abrantes et al,18 using Bangsbö’s repeated sprints test (ie, 6 × 34.2 m, departing every 25 s) that clearly discriminated Portuguese elite, regional adults, junior, and adolescent soccer players. In the only study that measured the ability to repeat explosive efforts, Sheppard et al4 showed that Australian national team players have higher and faster performances than players from the develop-ment national team.

In summary, the present findings support the general view that an athlete’s abil-ity to repeat explosive efforts is related to general fitness level, partly attributable to gender, natural ability, training history, and maturation process. Finally, performing repeated-shuttle runs might be considered as extreme changes of directions, when compared with those commonly observed in the field. Furthermore, performing countermovement jumps may not best represent the jumping patterns during games. However, the intension of this test is that it is generic and can be used to compare the athletic profiles of relevant team sports. Compared with a sport-specific test (eg, Sheppard et al4), the RSSJA test involves standard and common running distances, changes of directions, and jumping movements so that players from different team sports can easily cope with the protocol. While the present findings support the construct validity of the RSSJA test for handball players, investigating players from other team sports such as basketball or netball at different playing levels would elaborate on the construct validity of the test.

Practical Applications and Conclusions

The RSSJA test is reliable and valid to assess repeated explosive effort sequences (ie, repeated sprint and jump abilities simultaneously) in team sports such as handball. Owing to the low typical errors of measurement, the RSSJA test is appropriate to monitor changes in an athlete’s sprinting and jumping performance throughout a season. It is suggested that a performance improvement of at least 2.6% (–0.13 s) and 4.8% (+153 W) in mean sprint time and estimated jump peak power have to be detected to determine an “almost certain” increase in performance. In addition, since test results are likely to be representative of gender and playing level, the use of the test in talent identification purposes is also of interest.

Acknowledgments

The authors thank Irmant Cadjjiov for his assistance with manuscript preparation and the subjects for their enthusiastic participation. We also thank Pierre Mangin, head coach of the Women’s French national handball team, and Christine Renaud, Arnaud Parisy, and Francois Berthier for their support during the experimentations.


References 1. Spencer M, Bishop D, Dawson B, Goodman C. Physiological and metabolic responses of

repeated-sprint activities:specific to field-based team sports. Sports Med. 2005;35:1025–1044.

2. Luig P, Manchado Lopez C, Pers J, et al. Motion characteristics according to playing position in international men’s team handball. In 13th annual congress of the European college of sport science. Estoril, Portugal. 2008.

3. Ben Abdelkrim N, El Fazaa S, El Ati J. Time-motion analysis and physiological data of elite under-19-year-old basketball players during competition. Br J Sports Med. 2007;41:69–75, discussion 75.

4. Sheppard JM, Gabbett T, Taylor KL, Dorman J, Lebedew AJ, Borgeaud R. Develop-ment of a Repeated-Effort Test for Elite Men’s Volleyball. Int J Sports Physiol Perform. 2007;2:292–304.

5. Rampinini E, Bishop D, Marcora SM, Ferrari Bravo D, Sassi R, Impellizzeri FM. Validity of simple field tests as indicators of match-related physical performance in top-level professional soccer players. Int J Sports Med. 2007;28:228–235.

6. Impellizzeri FM, Rampinini E, Castagna C, et al. Validity of a Repeated-Sprint Test for Football. Int J Sports Med. 2008;29:899–905.

7. Wisloff U, Castagna C, Helgerud J, Jones R, Hoff J. Strong correlation of maximal squat strength with sprint performance and vertical jump height in elite soccer players. Br J Sports Med. 2004;38:285–288.

8. Mujika I, Santisteban J, Impellizzeri FM, Castagna C. Fitness determinants of success in men’s and women’s football. J Sports Sci. 2009;27:107–114.

9. Impellizzeri FM, Rampinini E, Castagna C, Bishop D, Ferrari Bravo D, Tibaudi A, and Wisloff U. Validity of a Repeated-Sprint Test for Football. Int J Sports Med. 2008;29:899–905.

10. Buchheit M, Millet GP, Parisy A, Pourchez S, Laursen PB, Ahmaidi S. Supramaximal training and post-exercise parasympathetic reactivation in adolescents. Med Sci Sports Exerc. 2008;40:362–371.

11. Spencer M, Fitzsimons M, Dawson B, Bishop D, Goodman C. Reliability of a repeated-sprint test for field-hockey. J Sci Med Sport. 2006;9:181–184.

12. Wragg CB, Maxwell NS, Doust JH. Evaluation of the reliability and validity of a soccer-specific field test of repeated sprint ability. Eur J Appl Physiol. 2000;83:77–83.

13. Cormack SJ, Newton RU, McGuigan MR, Doyle TL. Reliability of measures obtained during single and repeated countermovement jumps. Int J Sports Physiol Perform. 2008;3:131–144.

14. Clark RA. The effect of training status on inter-limb joint stiffness regulation during repeated maximal sprints. J Sci Med Sport. 2009;12:406–410.

15. Buchheit M, Leblond F, Renaud C, Kuhnle J, Ahmaidi S. Effect of complex vs. spe-cific aerobic training in young handball players. Coaching & Sport Science Journal. 2008;3:22.

16. Hopkins WG. How to interpret changes in an athletic performance test. Sportscience. 2004;8:1–7.

17. Weir JP. Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res. 2005;19:231–240.

18. Abrantes C, Maçãs V, Sampaio J. Variation in football players’ sprint test performance across different ages and levels of competition. J Sports Sci Med. 2004;3:44–49.

19. Leone M, Comtois AS. Validity and reliability of self-assessment of sexual maturity in elite adolescent athletes. J Sports Med Phys Fitness. 2007;47:361–365.

20. Glaister M, Howatson G, Lockey RA, Abraham CS, Goodwin JE, McInnes G. Famil-iarization and reliability of multiple sprint running performance indices. J Strength Cond Res. 2007;21:857–859.


21. Domire ZJ, Challis JH. The influence of squat depth on maximal vertical jump perfor-mance. J Sports Sci. 2007;25:193–200.

22. Sayers SP, Harackiewicz DV, Harman EA, Frykman PN, Rosenstein MT. Cross-validation of three jump power equations. Med Sci Sports Exerc. 1999;31:572–577.

23. Hopkins WG, Schabort EJ, Hawley JA. Reliability of power in physical performance tests. Sports Med. 2001;31:211–234.

24. Hopkins WG. Reliability from consecutive pairs of trials (Excel spreadsheet). A new view of statistics. sportsci.org: Internet Society for Sport Science, sportsci.org/resource/stats/xrely.xls. 2000.

25. Batterham AM, Hopkins WG. Making meaningful inferences about magnitudes. Int J Sports Physiol Perform. 2006;1:50–57.

26. Buchheit M, Laursen PB, Kuhnle J, Ruch D, Renaud C, Ahmaidi S. Game-based Train-ing in Young Elite Handball Players. Int J Sports Med. 2009;30:251–258.

27. Buchheit M. Field tests to monitor athletic performance throughout a team-sport season. Sci Sports. 2008;23:29–31.

28. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc. 2009;41:3–13.

29. Gorostiaga EM, Granados C, Ibanez J, Izquierdo M. Differences in physical fitness and throwing velocity among elite and amateur male handball players. Int J Sports Med. 2005;26:225–232.

30. Delextrat A, Cohen D. Physiological testing of basketball players: toward a standard evaluation of anaerobic fitness. J Strength Cond Res. 2008;22:1066–1072.

reliability, usefulness, and validity of a repeated sprint ... · pdf fileyoung elite women...

Documents