Journal of Science and Medicine in Sport (2008) 11, 202—208

ORIGINAL PAPER

The Yo—Yo intermittent recovery test inbasketball players

Carlo Castagnaa,b,∗, Franco M. Impellizzeri c, Ermanno Rampinini c,Stefano D’Ottaviob, Vincenzo Manzib

a Scuola Regionale dello Sport delle Marche, Italian Olympic Commitee (CONI), Italyb School of Sport and Exercise Sciences, University of Rome Tor Vergata, Italyc Human Performance Laboratory, S.S. MAPEI, Italy

Received 4 August 2006; received in revised form 23 February 2007; accepted 26 February 2007

KEYWORDSField testing;Shuttle running;Intermittent exercise;Line drill;Fatigue

Summary The purpose of this study was to examine the physiological correlatesof the Yo—Yo intermittent recovery test level 1 (Yo—Yo IR1) in basketball play-ers. Twenty-two male basketball players (means ± S.D., body mass 72.4 ± 11.4 kg,height 181.7 ± 6.9 cm, age 16.8 ± 2.0 years) were tested for maximal oxygen uptake(VO2max), ventilatory threshold (VT) and running economy (RE) on a motorized tread-mill. Lower limb explosive strength and anaerobic-capacity was assessed usingvertical jumps (CMJ), 15 m shuttle running sprint (15 mSR) and line drill (LD),respectively. The same test battery was replicated after an experimental basket-ball game in order to assess selective effect of fatigue on physical performance.Pre to post-game CMJ (40.3 ± 5.7 versus 39.9 ± 5.9 cm) and 15 mSR (5.80 ± 0.25versus 5.77 ± 0.22 s) performances were not significantly different (p > 0.05). LDperformance decreased significantly post-game (from 26.7 ± 1.3 to 27.7 ± 2.7 s,p < 0.001). Yo—Yo IR1 performances (m) were significantly related to VO2max (r = 0.77,

p = 0.0001), speed at VO2max (r = 0.71, p = 0.0001) and %VO2max at VT (r = −0.60,p = 0.04). Yo—Yo IR1 performance was significantly correlated to post-game LD decre-ments (r = −0.52, p = 0.02). These findings show that Yo—Yo IR1 may be consideredas a valid basketball-specific test for the assessment of aerobic fitness and game-related endurance.

Aus

d

© 2007 Sports Medicine

Introduction

Competitive basketball is an intermittent high-intensity physical activity that requires a well-

∗ Corresponding author.E-mail address: castagnac@libero.it (C. Castagna).

bdappt

1440-2440/$ — see front matter © 2007 Sports Medicine Australia. Publishedoi:10.1016/j.jsams.2007.02.013

tralia. Published by Elsevier Ltd. All rights reserved.

eveloped aerobic and anaerobic fitness.1 Althoughasketball performance is thought to be mainlyependent on players’ anaerobic ability, high

erobic fitness is also important for improvederformance.2,3 Specifically maximal aerobicower (VO2max) is considered to improve the abilityo recovery from the anaerobic efforts during the

d by Elsevier Ltd. All rights reserved.

F

gbel

rutcmaacopat((tcpv

mfYsaRissYchwvpiYb

itfitaYgbiimoY

sis

M

S

Tm1msasaiPfld

ttttpocia

E

IwiVortdbgncTpm(r

ield test in basketball players

ame.4 Furthermore, aerobic conditioning haseen suggested to be important for preparing play-rs to be able to sustain an appropriate trainingoad volume for basketball.2

Aerobic fitness (VO2max, lactate thresholds, andunning economy) can be accurately evaluatedsing a variety of laboratory protocols.5 Althoughhe values obtained with laboratory testing areonsidered the ‘‘gold standard’’ for the measure-ent of aerobic fitness, the procedures involved

re time consuming and require trained personnelnd expensive equipment. For these reasons, someontinuous field-tests6,7 involving shuttle runningver 20 m (approximately court length) have beenroposed as practical alternatives to laboratoryssessments in basketball.8 The validity of theseests is based on their correlations with VO2maxcriterion validity)6,7 and displacement specificitylogical validity). However, due to the intermit-ent nature of basketball,1 the exercise continuityonsidered in these tests6,7 may considered as aotential threat to the logical validity and contentalidity.9

Recently Bangsbo10 developed the Yo—Yo inter-ittent recovery test (Yo—Yo IR1) as a field test

or assessing performance for team-sport players.o—Yo IR1 consists of 2 m × 20 m bouts of progres-ive speed shuttle-running, interspersed by 10 s ofctive recovery, performed until exhaustion.10,11

ecent research has shown that Yo—Yo IR1 elic-ts maximal aerobic responses while significantlytressing the anaerobic energy system,11 demon-trating that the physiological demands of theo—Yo IR1 are similar to those taxed duringompetition.11 Furthermore, Yo—Yo IR1 scoresave been reported to be significantly correlatedith VO2max assessed in laboratory, confirming itsalidity as measures of aerobic-fitness in soccerlayers.12 Despite its logical validity and popular-ty no study has been performed in order to testo—Yo IR1 validity as measures of aerobic fitness inasketball players (population validity).9

The first aim of this study was to exam-ne the relationship between the Yo—Yo IR1 andraditional physiological parameters of aerobic-tness assessed in the laboratory, in a popula-ion of trained basketball players. The secondim was to examine the association betweeno—Yo IR1 performance and selected aspects ofame-performance1,13 in order to make possi-le statements of test specificity.14 Due to thentermittent high-intensity nature of Yo—Yo IR111

t was hypothesized that significant associationsay occur among maximal-intensity components

f aerobic-fitness and anaerobic-endurance witho—Yo IR1 performance.12 An improved under-

k

ds

203

tanding of these relationships may provide usefulnformation for fitness assessment and training pre-cription in basketball players.

aterial

ubjects

wenty-two male junior basketball players (bodyass 72.4 ± 11.4 kg, height 181.7 ± 6.9 cm, and age

6.8 ± 2 years) from the same basketball club (Sta-ura Basket, Ancona, Italy) were involved in the

tudy. All players had official medical clearanceccording to national law. Written informed con-ent was received from all players and parentsfter verbal and written explanation of the exper-mental design and potential risks of the study.articipants were told they were free to withdrawrom the study at any time without penalty. Theocal Institutional Review Board approved this studyesign.

The assessments were performed at the sameime of the day (from 5.00 to 7.00 p.m. for fieldests and from 4.00 to 8.00 p.m. for laboratoryests) and subjects were blinded about the aims ofhe study. All of the testing procedures were com-leted in June one week after the play-off phasef the basketball competitive season. During theompetitive season the training schedule consistedn five training sessions (∼90 min each session) withcompetitive game played during the weekend.

xperimental approach to the problem

n the present investigation, a descriptive designas used. Criterion validity was evaluated exam-

ning Yo—Yo IR1 performance relationship withO2max, ventilatory threshold (VT), running econ-my (RE) and speed attained at VO2max. Testelevance was also assessed by examining the rela-ionship between Yo—Yo IR1 performance and scoreecrements of basketball-specific tests performedefore and at the end of an experimental basketballame.13 It was assumed that the occurrence of sig-ificant pre-to-post game test variations8,15,16 wereonsequence of selective game related fatigue.13

o achieve this, other common basketball-specificerformance tests (i.e. the vertical counter-ovement jump (CMJ), 15 m shuttle running sprint

15 mSR) and line drill (LD)), were completed in aandom order before and after a competitive bas-

etball match.

Testing sessions took place over three separateays (at least 72 h a part). The testing sessions con-isted as follows:

riwib8ais1pwgRa

bYucVte6cret

E

204

1. Treadmill-testing;2. Yo—Yo IR1;3. Experimental game.

Yo—Yo intermittent recovery test

Yo—Yo intermittent recovery test features two lev-els differing with respect to starting speed.10 In thisinvestigation, the level 1 version was completedaccording to previously described methods.11 Allplayers were familiar with the testing proceduresbeing part of their usual fitness assessment pro-gram. Yo—Yo IR1 test consisted of 20-m shuttleruns performed at increasing velocities (Table 1)with 10 s of active recovery between runs untilexhaustion. The end of the test was consideredwhen the participant twice failed to reach thefront line in time (objective evaluation) or he feltnot able to cover another shuttle at the dictatedspeed (subjective evaluation). The total distancecovered during the Yo—Yo IR1 was primary perfor-mance measure10 and the speed attained during thelast 2 m × 20 m bout was considered as Vmax.12 Allfield testing sessions were performed on the samebasketball court where players usually trained.Reliability of the Yo—Yo test calculated as a CVis 4.9% for the total distance covered during thetest.11

Aerobic fitness assessment

Maximum oxygen uptake was determined using anincremental running test on a motorized treadmill(RunRace, Technogym, Gambettola, Italy). Afteran individually adjusted warm-up (5 min) subjects

fi

Cfi

Table 1 Yo—Yo intermittent recovery (level 1) test protoco

Stage Speed (km h−1) Shuttle bouts (2 m × 20

1 10 12 12 13 13 24 13.5 35 14 46 14.5 87 15 88 15.5 89 16 8

10 16.5 811 17 812 17.5 813 18 814 18.5 815 19 8

After each 2 m × 20 m shuttle-run bout players perform 10 s of acstarting line.

C. Castagna et al.

un for 6 min at 8 km h−1, and the velocity wasncreased by 1 km h−1 every minute until 16 km h−1

as attained. Thereafter exercise intensity wasncreased through increasing the treadmill gradey 1◦ min−1 until exhaustion, that was reached in—12 min. Achievement of VO2max was considereds the attainment of at least two of the follow-ng criteria: (1) a plateau in VO2 despite increasingpeeds, (2) a respiratory exchange ratio above.10, (3) a heart rate (HR) ± 10 beats min−1 of age-redicted maximal HR (220-age). Expired gasesere analyzed using a breath-by-breath automatedas-analysis system (K4b2, COSMED, Rome, Italy).eliability, accuracy and validity of the K4b2 gasnalyzer system has been reported elsewhere.17,18

Before each test flow and volume were cali-rated using a 3-L capacity syringe (Sensormedics,orba Linda, CA). Gas analyzers were calibratedsing gases of (O2 and carbon dioxide) knownoncentrations (Sensormedics, Yorba Linda, CA).entilatory threshold (VT) was assessed accordingo Beaver et al.19 Running economy was consid-red as average VO2 during the final minute of themin run at 8 km h-1. Maximal aerobic speed wasalculated using the relationship between VO2 andunning speed.20 Maximal HR (HRmax) was consid-red as the highest 5 s mean during the treadmillest.

xplosive-strength and anaerobic-capacity

eld tests

ountermovement jump,21 15 mSR and LD per-ormance were measured 20 min before andmmediately after the experimental game in ran-

l

m) Split distance Accumulated distance

40 4040 8080 160

120 280160 440320 760320 1080320 1400320 1720320 2040320 2360320 2680320 3000320 3320320 3640

tive recovery jogging round a cone set 5 m apart from the

F

d1sCNt(cthmsiiac

mbPwsbcaarwma(11

E

Tes3itdtoott(Fu(lfdJ

S

DtaScuYspuYVaUeSStU

R

Vist(caeYp

are shown in Table 3. During the first and the secondhalf of the experimental game, players attained86 ± 5 and 87 ± 4% of the individual HRmax, respec-tively (p = 0.42). Experimental-games blood lactate

Table 2 Physiological characteristics of the basket-ball players (n = 22)

Variables Mean ± S.D.

VO2max (mL kg−1 min−1) 60.4 ± 5.1VT VO2 (mL kg−1 min−1) 40.2 ± 4.7RE (mL kg−1 min−1) 31.5 ± 4.3Yo—Yo IR1 distance (m) 1678 ± 397Yo—Yo IR1 velocity (km h−1) 15.7 ± 0.7Speed at VO2max (km h−1) 17.1 ± 1.8vVT (km h−1) 11.6 ± 1.3

ield test in basketball players

omized order. Pre-test warm-up consisted in0 min basketball skills practice conducted atelf selected intensity followed by sub-maximalMJs (2—3 repetitions) and shuttle-run sprints.o stretching exercises were allowed prior tohe test. Countermovement jump performancejump height) was measured using a switch-matonnected to a computer (Muscle Lab, Bosco Sys-em, Rieti, Italy). Subjects were asked to keepands on their hips to prevent influence of armovements on vertical jump performance.21 Each

ubject performed at least two maximal CMJs start-ng from a standing position, with 1 min recoveryn-between. Players were asked to jump as highs possible. The best jump height was used foralculations.

Fifteen-meter shuttle running sprint perfor-ance was assessed with the aid of a photocelleam connected to a personal computer (Microgateolifemo, Bolzano, Italy). Photocell beam heightas set at 45 cm from ground level with players

tarting from a line set 50 cm from the photocelleam. The best of two 15 mSR trials was used foralculation (2 min passive recovery between tri-ls). The LD15,22 performance was measured usinghand-held stopwatch (Casio, HS-30W, Japan). The

eliability of CMJ, 15 mSR and LD (reported as CV%)ere assessed during the week before the com-encement of the study in the same players (n = 22)

nd ranged between 0.5 and 0.9%. The variationsCV%) within the measurements trials for CMJ and5 mSR on the experimental day were less than%.

xperimental basketball-game

wenty players performed two basketball gamesach lasting two match play periods of 10 min inter-persed by a 2 min interval (mean total play time0.5 min). This time duration was chosen as thiss the average playing time12 for the competi-ive level of the players in this study (unpublishedata). During the experimental games no substi-ution was allowed in order to have all playersbserved for the same amount of playing time andnly man to man defense was allowed. In ordero evaluate exercise intensity HR was monitoredhroughout the game with a HR monitor systemPolar Team System, Polar Electro Oy, Kempele,inland). Post-hoc HR analyses were performedsing Polar Precision System SW 3.0 softwarePolar Electro Oy, Finland). Furthermore blood

actate concentration measurements were per-ormed sampling players’ ear lobe blood randomlyuring the matches (Lactate Pro, Arkray, Tokyo,apan).

205

tatistical analyses

ata are reported as mean ± standard devia-ion (S.D.). Before using parametric tests, thessumption of normality was verified using thehapiro—Wilk W test. Pearson’s product—momentorrelations and linear regression analysis weresed to examine the relationships between theo—Yo IR1 performance (m) and VO2max, VTpeed, RE, speed at VO2max and LD post-gameerformance decrement. One-way ANOVA wassed to examine possible differences betweeno—Yo IR1 maximal speed, speed at VO2max andT speed. Comparisons between field-test vari-ble means was performed using paired t-tests.npaired t-tests were performed to examine play-rs’ game-role differences over test variables.ignificance was assumed at 5% (p ≤ 0.05) a priori.tatistical analyses were performed using Statis-ica package (6.0 version, Statsoft, California,SA).

esults

alues of the aerobic-fitness variables consideredn this study are shown in Table 2. Speed at VT,peed at VO2max and Yo—Yo IR1 speed at exhaus-ion were significantly different from each otherp = 0.002). Yo—Yo IR1 performance was signifi-antly correlated with VO2max (r = 0.77, p = 0.0001)nd speed at VO2max (r = 0.71, p = 0.0001). How-ver, no significant correlation was found betweeno—Yo IR1 and VT speed and RE (r = −0.12,> 0.05).Results of the pre-post experimental-game tests

HRmax (beats min−1) 198 ± 6

RE: VO2 at 8 km h−1, VT VO2: VO2 at ventilatory threshold;vVT: speed at ventilatory threshold, HRmax: maximal heartrate.

206

Table 3 Pre-post experimental-game test perfor-mances (n = 20)

Variables Pre-game Post-game

CMJ height (cm) 39.9 ± 5.9 40.3 ± 5.715 mSR 5.80 ± 0.25 5.77 ± 0.22Line drill 26.7 ± 1.3 27.7 ± 2.7a

a Significance at p ≤ 0.001.

D

Tobbibv

frse(5aac

cwTbfojpppmY

Figure 1 Scatter plot of VO2max (mL kg−1 min−1) vs.Yo—Yo IR1 test performance (m).

concentration was 3.7 ± 1.4 mmol l−1. Players weresampled for blood lactate concentration a minimumof two times per match (45 blood samplings).

Significant correlations were found between pre-post LD decrements and Yo—Yo IR1 performance(r = −0.52, p = 0.02). No significant correlations

were observed between the Yo—Yo IR1 test perfor-mance and game mean HR or first versus second halfHR variations (Figs. 1 and 2).

Figure 2 Scatter plot of line drill decrements (%) vs.Yo—Yo IR1 test performances (m).

slpiYreifaaaaiip

roatsp

C. Castagna et al.

iscussion

his is the first study that demonstrates the validityf Yo—Yo IR1 test as measure of aerobic-fitness inasketball players and its specificity as measure ofasketball-specific endurance. This is of particularnterest for sport scientists as Yo—Yo IR1 test haseen proposed as suitable for basketball players butalidated in soccer players only.12

This study showed that Yo—Yo IR1 test per-ormance and VO2max values were significantlyelated, with 59% of shared variance. This isimilar to previous research that reported a mod-rate correlation (r = 0.71, p < 0.05) in a groupn = 15) of habitually active subjects (VO2max0.5 mL kg−1 min−1).11 Although the strength of thisssociation confirms the validity of Yo—Yo IR1 tests generic measure of aerobic fitness, it is not suffi-iently high to demonstrate its predictive validity.9

The present results revealed that of the plannedomparisons only VO2max and speed at VO2maxere associated with Yo—Yo IR1 test performance.herefore the finding of a significant relationshipetween speed at VO2max and Yo—Yo IR1 test per-ormance suggests that this test is influenced byther components of aerobic-performance thanust VO2max.20,23,24 Moreover, the Yo—Yo IR1 testerformance was also inversely related to theercentage of VO2max attained at VT (r = −0.60,= 0.04), partially supporting the hypothesis of aore complex influence of aerobic-components to

o—Yo IR1 test performance.The Yo—Yo IR1 performance observed in this

ample of basketball players (1678 ± 397 m) wasower than those observed in professional soccerlayers at the beginning (1760 ± 59 m) and dur-ng the competitive season (2211 ± 70 m).11 Highero—Yo IR1 performances were also reported inegional level soccer players (2138 ± 364).12 How-ver, the mean Yo—Yo IR1 performance foundn this study was higher than that reportedor female top-level soccer players (1379 m)25

nd for state level (840—1049 m) and recre-tional level (1010—1048 m) Australian team-sportthletes.26 Unfortunately, to the best of this studyuthor’s knowledge no information is availablen the international scientific literature regard-ng Yo—Yo IR1 test performance in basketballlayers.

Differences in Yo—Yo IR1 performance have beeneported to be consequence of fitness status, periodf the competitive season, and playing role within

team.11 In this study players were tested after

he completion of a successful competitive sea-on and as consequence of that possessing a goodhysical fitness. Indeed, the basketball players

F

ste(weir2ICsp(mtrb

owmGnppleu

uogwYbtidtLmascvttbsi

tig1

tqoeoIwvhptaip

rasnsHdetdtbmttbogatloiksYsumttgim

tm

ield test in basketball players

howed a VO2max (60.4 ± 5.1 mL kg−1 min−1) higherhan what usually reported for basketball play-rs competing at elite level.27 Furthermore, CMJ40.3 ± 5.7 cm) and LD (26.7 ± 1.3 s) performancesere similar or even higher to what reported inlite basketball teams of the same age. Specif-cally, Apostolidis et al.27 and Hoffman et al.15

eported LD performances of 27.92 ± 1.04 s and8.3 ± 0.9 s for junior players of the Greek andsraeli National teams, respectively. Moreover, theMJ performance of the players in this study wereimilar to those reported in professional adultlayers (43.9 ± 4.0)28 and in junior elite players40.1 ± 3.7 cm).27 Given the good physical perfor-ances showed by this sample of basketball players

he Yo—Yo IR1 test performance here reported mayepresent a valid reference for competitive basket-all.

In order to assess possible role dependent effectsver Yo—Yo IR1 test performance11,29 playersere grouped according to main playing assign-ents in two groups: Forwards (Centers-Wings anduards-Playmakers, n = 11 each). Results showedo significant positional effects on Yo—Yo IR1 testerformance, pre to post match performances andhysiological variables. These findings are mostikely due to homogeneity of the performance lev-ls and/or training status of the basketball playerssed in this study.

A field test cannot be considered sport-specificntil a direct association between some aspectsf match performance and the field test investi-ated has been determined.14 In line with this,e attempted to assess the specificity of the

o—Yo IR1 test by examining the relationshipsetween the changes in basketball-specific fieldests undertaken before and just after an exper-mental basketball-game with distance covereduring the Yo—Yo IR1 test.13,30,31 We assumedhat variations in test scores (CMJ, 15 mSR andD) may directly reflect the fatigue effects ofatch play.13,30,31 Due to this, it was imper-

tive that the game intensity obtained in thistudy was specific to that encountered in actualompetitive basketball match play. Although obser-ations were made using an experimental-game,he intensity attained by players during competi-ion are similar to game intensities in professionalasketball games.1 These results provide furtherupport the validity of the research design in thisnvestigation.

The preplanned correlation analysis revealed

hat Yo—Yo IR1 test performance was only signif-cantly related to the pre to post experimental-ame LD decrements, but not the changes in the5 mSR and CMJ. The finding that explosive short

bYtb

207

erm maximal efforts (<6 s) did not vary as conse-uence of match play agrees with previous researchn competitive soccer players.13,31 In the Hoffmant al.13 study no changes in CMJ performance werebserved following a competitive soccer match.t was speculated that CMJ performance (Power)as maintained through the match because of ele-ated catecholamine levels during match play thatave been shown to positively affect anaerobicower development.32 Conservation of short-erm maximal performance has been reportedlso by others who demonstrated no changesn isokinetic strength following soccer matchlay.33,34

More recently Krustrup et al.31 found thatepeated sprint performance (5 m × 30 m with 25 sctive recovery) changed following experimentaloccer-match. In particular these authors reportedo changes in the initial sprint of the repeatedprint bout during the first half of the match.owever, sprint bout performance was found toecrease following the third sprint after an intensexercise period in the first half. In contrast,o the first half measures, sprint performanceecreased from the first sprint bout compared tohe control condition. In the present study, theasketball players reported a significant decre-ent in LD after the experimental game, showing

hat basketball playing may induce fatigue pat-erns (repeated sprint) similar to those inducedy a soccer match just after approximately 20 minf match play (∼30 min of total playing). We sug-est that the relationship between Yo—Yo IR1 testnd game related LD decrements may be dueo selective glycogen depletion patterns in lowerimbs fast-twitch fibers.11,31 Combined with previ-us research,31 the present results provide usefulnformation that can be used to develop bas-etball specific training strategies. Moreover, theignificant although moderate association betweeno—Yo IR1 test performance and LD decrementsuggest that the Yo—Yo IR1 test may be aseful field test for assessing endurance perfor-ance in basketball players. However in order

o fully assess test ecological validity, the rela-ionship between Yo—Yo IR1 test and selectedame-activities should be examined through phys-cal performance assessment within a competitiveatch.This study is the first to report on the validity of

he Yo—Yo IR1 test and aerobic fitness measures andatch-related performance decrements in basket-

all players. The present findings suggest that theo—Yo IR1 test may be considered a valid field-testo assess aerobic-fitness and game preparedness ofasketball players.

208

Practical implications

• The Yo—Yo intermittent recovery test (level 1)is a valid test to assess aerobic fitness in maleyoung basketball players.

• The ability to perform short-term (∼30 s)maximal-intensity exercise decreases as con-sequence of basketball playing.

• The Yo—Yo intermittent recovery test (level 1)performance is related to post-game decre-ment in short-term (∼30 s) basketball-specificmaximal anaerobic-performance.

References

1. McInnes SE, Carlson JS, Jones CJ, et al. The physiologicalload imposed upon basketball players during competition.J Sports Sci 1995;13:387—97.

2. Stone WJ, Steingard PM. Year-round conditioning for bas-ketball. Clin Sports Med 1993;12(2):173—91.

3. Stone WJ, Kroll WA. Sports conditioning and weight train-ing. 3rd ed. Dubuque, Iowa: WmC Brown Publishers; 1991.

4. Tomlin DL, Wenger HA. The relationship between aerobicfitness and recovery from high intensity intermittent exer-cise. Sports Med 2001;31(1):1—11.

5. Pate RR, Kriska A. Physiological basis of the sex differencein cardiorespiratory endurance. Sports Med 1984;1(2):87—98.

6. Ramsbottom R, Brewer J, Williams C. A progressive shuttlerun test to estimate maximal oxygen uptake. Br J SportsMed 1988;22(4):141—4.

7. Leger L, Gadoury C. Validity of the 20 m shuttle run testwith 1 min stages to predict VO2max in adults. Can J SportSci 1989;14(1):21—6.

8. Stapff A. Protocols for the physiological assessment of bas-ketball players. In: Physiological tests for elite athletes.Champaign, IL: Human Kinetics Publishers; 2000.

9. Thomas JR, Nelson JK, Silverman J. Research methods inphysical activity. 5th ed. Champaign, IL: Human Kinetics;2005.

10. Bangsbo J. Yo-yo test. Ancona: Kells; 1996.11. Krustrup P, Mohr M, Amstrup T, et al. The Yo—Yo intermit-

tent recovery test: Physiological response, reliability, andvalidity. Med Sci Sports Exerc 2003;35(4):697—705.

12. Castagna C, Impellizzeri FM, Chamari K, et al. Aerobicfitness and yo-yo continuous and intermittent tests perfor-mances in soccer players: a correlation study. J StrengthCond Res 2006;20(2):320—5.

13. Hoffman JR, Nusse V, Kang J. The effect of an intercolle-

giate soccer game on maximal power performance. Can JAppl Physiol 2003;28(6):807—17.

14. Boddington MK, Lambert MI, Waldeck MR. Validity of a 5-mmultiple shuttle run test for assessing fitness of women fieldhockey players. J Strength Cond Res 2004;18(1):97—100.

Available online at www.s

C. Castagna et al.

15. Hoffman JR, Epstein S, Einbinder M, et al. A Comparisonbetween the wingate anaerobic power test to both verticaljump and line drill tests in basketball players. J StrengthCond Res 2000;14(3):261—4.

16. Hoffman JR, Stavsky H, Falk B. The effect of water restric-tion on anaerobic power and vertical jumping height inbasketball players. Int J Sports Med 1995;16:214—8.

17. McLaughlin JE, King GA, Howley ET, et al. Validation of theCOSMED K4 b2 portable metabolic system. Int J Sports Med2001;22(4):280—4.

18. Duffield R, Dawson B, Pinnington HC, et al. Accuracy andreliability of a Cosmed K4b2 portable gas analysis system.J Sci Med Sport 2004;7(1):11—22.

19. Beaver WL, Wasserman K, Whipp BJ. A new method fordetecting anaerobic threshold by gas exchange. J ApplPhysiol 1986;60(6):2020—7.

20. Billat V, Koralsztein JP. Significance of the velocity atVO2max and time to exhaustion at this velocity. Sports Med1996;22:90—180.

21. Bosco C, Luhtanen P, Komi PV. A simple method for measure-ment of mechanical power in jumping. Eur J Appl Physiol1983;50:273—82.

22. Semenick D. Tests and measurements: The line drill test.NSCA J 1990;12(2):47—9.

23. Morgan DW, Baldini FD, Martin PE, et al. Ten kilo-meter performance and predicted velocity at VO2maxamong well-trained male runners. Med Sci Sports Exerc1989;21(1):78—83.

24. Noakes TD, Myburgh KH, Schall R. Peak treadmill runningvelocity during the VO2max test predicts running perfor-mance. J Sports Sci 1990;8(1):35—45.

25. Krustrup P, Mohr M, Ellingsgaard H, et al. Physical demandsduring an elite female soccer game: importance of trainingstatus. Med Sci Sports Exerc 2005;37(7):1242—8.

26. Thomas A, Dawson B, Goodman C. The Yo—Yo test: reliabil-ity and association with a 20-m shuttle run and VO2max. IntJ Sports Physiol Perform 2006;1:137—49.

27. Apostolidis N, Nassis GP, Bolatoglou T, et al. Physiologi-cal and technical characteristics of elite young basketballplayers. J Sports Med Phys Fitness 2004;44(2):157—63.

28. Hakkinen K. Effects of the competitive season on physicalfitness profile in elite basketball players. J Human Mov Stud1988;15:119—28.

29. Mohr M, Krustrup P, Bangsbo J. Match performance ofhigh-standard soccer players with special reference todevelopment of fatigue. J Sports Sci 2003;21:519—28.

30. Bangsbo J. The physiology of soccer - with special refer-ence to intense intermittent exercise. Acta Physiol Scand1994;151(Suppl. 619):1—155.

31. Krustrup P, Mohr M, Steensberg A, et al. Muscle and bloodmetabolites during a soccer game: implications for sprintperformance. Med Sci Sports Exerc 2006;38(6):1165—74.

32. Kraemer WJ, Patton JF, Knuttgen H, et al. Effects of high-intensity cycle exercise on sympathoadrenal-medullary

response patterns. J Appl Physiol 1991;70(1):8—14.

33. Ekblom B. Applied physiology of soccer. Sports Med1986;3:50—60.

34. Kirkendall DT. Physiology of soccer. In: Exercise and sportscience. Philadelphia: Lippincott Williams & Wilkins; 2000.

ciencedirect.com