1

1 Title: Influence of Oxygen Uptake Kinetics on Physical

2 Performance in Youth Soccer

4 Authors: Greg Doncaster1, 2, Simon Marwood3, John Iga4, 5 & Viswanath Unnithan1

1
6 Institutional Affiliations: Staffordshire University (UK),
2
7 Edge Hill University (UK),
3
8 Liverpool Hope University (UK),
4
9 Liverpool John Moores University (UK),
5
10 The Football Association (UK)

11

12 Corresponding Author: Viswanath Unnithan,

13 Professor of Paediatric Exercise Physiology,

14 Faculty of Health Sciences,

15 Staffordshire University,

16 ST4 2DF,

17 01782 295986,

18 [email protected]

19

20 Acknowledgements

21 We would like to thank all the participants involved in the study for their committed participation. The

22 authors would also like to thank the technical staff in the Faculty of Health Sciences at Staffordshire

23 University for their excellent support.

24 Finally, the authors are extremely grateful to Professor Barry Drust for his expert advice and support.

1
 2

25 Abstract

26 Purpose: To examine the relationship between oxygen uptake kinetics (VO 2 kinetics) and physical measures

27 associated with soccer match-play, within a group of highly trained youth soccer players.

28 Methods: Seventeen highly trained youth soccer players (age: 13.3 ± 0.4 yr, self-assessed Tanner Stage: 3 ± 1)

29 volunteered for the study. Players initially completed an incremental treadmill protocol to exhaustion, to establish

30 gaseous exchange threshold (GET) and VO 2max (59.1 ± 5.4 mL.kg-1.min-1). On subsequent visits players completed

31 a step transition protocol from, rest–moderate intensity exercise, followed by an immediate transition from

32 moderate–severe intensity exercise (moderate: 95%GET, severe: 60%∆), during which VO 2 kinetics were

33 determined. Physical soccer-based performance was assessed using a maximal Yo-Yo Intermittent Recovery test

34 level 1 (Yo-Yo IR1) and via GPS derived measures of physical soccer performance during soccer match-play, 3

35 2 x 20min, 11 v 11 matches, to gain measures of physical performance during soccer match-play.

36 Results: Partial correlations revealed significant inverse relationships between the unloaded to moderate transition

37 time constant (tau) and: Yo-Yo IR1 performance (r = -0.58, P = 0.02) and GPS variables (Total distance (TD): r

38 = - 0.64, P = 0.007, High speed running (HSR): r = - 0.64, P = 0.008, High speed running efforts (HSReff): r = -

39 0.66, P = 0.005).

40 Conclusion: Measures of VO 2 kinetics are related to physical measures associated with soccer match-play and

41 could potentially be used to distinguish between those of superior physical performance, within a group of highly-

42 trained youth soccer players.

43

44

45

46

47 Key Words: Team Sports, Intermittent Exercise, Youth Soccer, VO 2 Kinetics

48

49

50

51

52

2
 3

Abbreviations:

GET Gaseous exchange threshold

GPS Global positioning system

HSR High speed running

HSReff High speed running efforts

RER Respiratory exchange ratio

tau Time constant

TD Total distance

VHSR Very high speed running

VHSReff Very high speed running efforts

VO 2 kinetics Oxygen uptake kinetics

VO 2max Maximal oxygen consumption

Yo-Yo IR1 Yo-Yo intermittent recovery level 1 test

53

54

55 Introduction

56 The contribution of aerobic metabolism during soccer match play is dominant, accounting for up to 90% of the

57 energy required (Iaia, Rampinini & Bangsbo, 2009). However, as a consequence of the numerous, discrete high

58 intensity actions incurred throughout competitive match-play (Bangsbo, Mohr & Krustrup, 2006), the contribution

59 of anaerobic energy pathways is likely to be high at certain points during the game. The ability to perform, repeat

60 and sustain high intensity exercise is essential for elite soccer performance (Iaia et al., 2009). Consequently, faster

61 oxygen uptake kinetics (VO 2 kinetics) during transitions from low to high intensity exercise could preserve the

62 capacity of the anaerobic energy systems and thus enhance an individual’s tolerance to fatigue (Glaister, 2005).

63 In addition, faster VO 2 kinetics could result in a higher total energy turnover during high intensity exercise, when

64 added to maximal anaerobic energy turnover. Therefore, resulting in the ability to maintain, repeat and sustain

65 high intensity exercise during soccer match-play.

66

67 Global measures of aerobic capacity, like maximal oxygen consumption (VO 2max ), have been examined in relation

68 to measures of physical soccer performance (Rebelo, Brito, Seabra, Oliveira & Krustrup, 2014). Measures of

3
 4

69 VO 2max , however, have shown an inability to distinguish between players of the same playing level (Rampinini et

70 al., 2010; Stroyer, Hansen & Klausen, 2004) and is less sensitive to training when compared to the Yo-Yo IR1 &

71 IR2 tests, which are measures of intermittent running capacity (Bangsbo, Iaia & Krustrup, 2008). Measures of

72 VO 2 kinetics, therefore have been proposed as an alternative determinant of sports performance (Burnley & Jones,

73 2007). For example, while Rampinini et al (2010) reported no significant differences in measures of VO 2max

74 between professional and amateur adult soccer players, the professional soccer players were shown to possess

75 faster VO 2 kinetics than their amateur counter-parts, as measured by the time constant of the transition to moderate

76 intensity exercise (27.2 ± 3.3 vs. 32.3 ± 5.7 s). Furthermore, Rampinini et al. (2010) demonstrated that the time

77 constant from rest to moderate intensity exercise, was ‘largely’ related to both the Yo-Yo IR1 (r = -0.60) and Yo-

78 Yo IR2 (r = -0.65). Whether this effect has an impact on soccer match-play or is apparent within a cohort of highly

79 trained youth soccer players is yet to be investigated.

80

81 Research within adult populations has produced equivocal findings, with some studies demonstrating significant

82 associations between superior VO 2 kinetics and intermittent exercise capacity (Dupont et al., 2005; Rampinini et

83 al., 2009; Rampinini et al., 2010), while others have reported no association between the two (Buchheit, Hader &

84 Mendez-Villanueva, 2012; Christensen et al., 2011; Wells et al., 2012). Research in both male and female

85 adolescents, however, has demonstrated faster VO 2 kinetics in soccer trained individuals in comparison to their

86 untrained counterparts (Marwood et al., 2011; Unnithan, Roche, Garrard, Holloway & Marwood, 2015).

87 Consequently, VO 2 kinetics may impact upon measures of physical soccer performance during elite youth soccer

88 match-play. Previous studies, however, have limited applicability to performance during soccer match-play as

89 they only compared untrained and trained participants VO 2 kinetics during laboratory controlled treadmill exercise

90 and did not relate this to performance during soccer match-play. As a result, while soccer trained individuals have

91 been shown to demonstrate superior VO 2 kinetics when compared to untrained or amateur athletes, it is unclear

92 to what extent measures of VO 2 kinetics are related to physical measures of performance, obtained during soccer

93 match-play, within a group of highly-trained adolescent soccer players.

94

95 Initial research examining VO 2 kinetics in soccer have assessed VO 2 kinetics using a single square-wave transition

96 (unloaded – moderate intensity) protocol (Rampinini et al., 2010; Marwood et al., 2011; Unnithan et al., 2015).

97 Soccer match-play, however, incorporates severe intensity exercise which is often initiated from an elevated

4
 5

98 baseline work rate (VO 2 ), resulting in the recruitment of higher order motor units (preferential recruitment of type

99 II muscle fibres) (Nyberg et al., 2016). Consequently, in order to mimic these soccer-specific workload transitions,

100 a novel methodological approach was used by Nyberg et al. (2016). These authors utilised a double-step exercise

101 protocol, when examining the impact of additional speed endurance training on Yo-Yo IR1 performance, in highly

102 trained adult soccer players. The examination of VO 2 kinetics profiles during a step transition, work-to-work

103 protocol (unloaded – moderate and moderate - severe) therefore allows for a more representative test in a soccer-

104 specific cohort and provides a novel means for examining VO 2 kinetics in relation to the demands that are

105 presented during soccer match play. As a result, the purpose of this study was to examine the relationship between

106 VO 2 kinetics, during a step transition work-to-work protocol, and measures of physical performance obtained

107 during soccer match-play, within a group of highly trained youth soccer players. It was hypothesised that those

108 players who demonstrate faster pulmonary VO 2 kinetics during both transitions (unloaded – moderate & moderate

109 – severe) would also demonstrate superior measures of physical performance during soccer match-play.

110

111 Methods

112 Participants

113 Seventeen highly trained youth soccer players aged between 12 and 14 years volunteered to participate in this

114 study. All participants were outfield players (5 defenders, 6 midfielders and 6 attackers) from the same Category

115 One Premier League Football Academy. Prior to the commencement of the study, all players completed medical

116 health questionnaires and training history questionnaires. Table 1 displays all anthropometric and screening

117 measures. Maturation status was quantified using self-assessment, Tanner Stage method (Tanner, 1962) and

118 maturity offset (Mirwald, Baxter-Jones, Bailey & Beunen, 2002). Players and their parents were informed about

119 all procedures and requirements involved before providing written informed consent and assent from parents and

120 participants, respectively. Ethical approval was granted from the local university ethics committee.

121

122 *** Insert Table 1 Here***

123

124 Study Design

125 Data collection was conducted over a three-week period at the beginning of a competitive season (pre-season) at

126 a period where there were no competitive matches; during this period laboratory and field tests were conducted.

5
 6

127 The players visited the laboratory on 3 separate occasions, performed a field test (Yo-Yo IR1) and were involved

128 in three 11 v 11 matches (including the same pitch dimensions, the same composition of players in the same

129 positions and the same coaches, excluding goalkeepers from the analysis). During the periods of laboratory testing,

130 training only consisted of light technical practices, however, recovery periods of 24 hours were implemented

131 between all laboratory tests. In the first laboratory visit, players performed an incremental, ramp treadmill protocol

132 for the assessment of players’ gaseous exchange threshold (GET) and VO 2max . Players returned to the laboratory

133 on two further occasions. These two, separate visits to the laboratory were identical and required the participant

134 to complete a step transition protocol (from rest - moderate exercise, followed by an immediate transition from

135 moderate - severe intensity exercise) on a motorised treadmill for the assessment of their pulmonary oxygen

136 uptake kinetics. All testing was completed at the same time of day (± 2 hours), with room temperature, humidity

137 and pressure corresponding to 20.5 ± 1.5 ºC, 61.0 ± 1.4 % and 1016 ± 3 mmHg respectively, for laboratory testing.

138

139 Once all laboratory testing had been completed and a minimum of 48 h recovery had been undertaken, players

140 performed a maximal Yo-Yo IR1 field test. Subsequent to this and following a minimum of 48 h recovery, the

141 players’ physical performance during soccer match play was assessed during 3 separate, 11 v 11 matches, with a

142 minimum of 48 h recovery between each match. Matches were comprised of 2 x 20 min halves, with a 5 min rest

143 interval in between halves. Players’ match activities were monitored and analysed using 10 Hz global positioning

144 systems (GPS; Catapult, Melbourne, Australia).

145

146 All testing procedures (laboratory & field) were preceded by a 10 min warm-up, consisting of low intensity

147 running, dynamic stretching and then moderate intensity running. Following all tests a 5 min cool down, consisting

148 of low intensity running and static stretching, was conducted. All field testing and matches were conducted on

149 third generation artificial pitch in clear and dry conditions with minimal wind. Mean temperature, humidity and

150 pressure were 19.8 ± 2.4 ºC, 59.0 ± 3.4 % and 1009 ± 1 mmHg, respectively. Participants were instructed to refrain

151 from exercise on the days preceding each test and to maintain a normal diet throughout testing. Players were also

152 informed to refrain from consuming any drinks containing sugar or caffeine as well as the consumption of any

153 food in the two hours preceding any test.

154

155

156

6
 7

157 Assessment of Maximal Oxygen Uptake and Gaseous Exchange Threshold

158 Upon arrival at the laboratory and following pre-participation screening procedures (including completion of

159 training and health questionnaires as well as the assessment of players’ resting heart rate and blood pressure),

160 participants were fitted with a Polar Heart rate monitor (Polar Electro, Kempele, Finland) and face-mask (Hans

161 Rudolph, Hans Rudolph, Kansas City, USA), which was connected to an online gas analysis system (Cortex

162 MetaMax 3B, Cortex Biophysik GmbH, Leipzig, Germany). The online gas analyser was calibrated prior to each

163 visit according to the manufacturer’s instructions, using a known gas concentration and a 3-L syringe for manual

164 volume calibration of the ventilation sensors. Following a standardised 10 min warm up and a full description of

165 the test and safety procedures, participants began to run at a speed of 8 km/h at a 1 % incline (Jones & Doust,

166 1996) on a motorised treadmill (HP Cosmos, Pulsar, Sportgerate GmbH, Nussdorf, Germany). The speed of the

167 treadmill was increased by 1 km/h-1 every two minutes, this continued until participants reached 90% of their age

168 predicted heart rate max (207 – (0.7 x age)) (Gellish et al., 2007). At this point, the treadmill speed remained

169 constant whilst the incline of the treadmill was increased by 1% every minute until volitional exhaustion; this

170 procedure was employed to avoid over-striding and potential early termination and inaccurate assessment of

171 participant’s VO 2max . This method has been successfully used before to elicit VO 2max during treadmill running in

172 young populations (Williams, Carter, Jones & Doust, 2001). VO 2max was taken as the highest 15 sec average

173 during the test (Astorino, Robergs, Ghiasvand, Marks & Burns, 2000). Test criteria for a maximal effort, were the

174 achievement of at least two of the following three performance values: 1) respiratory exchange ratio (RER) above

175 1.00, 2) plateau in VO 2 despite an increasing speed (increase in VO 2 of < 2 mL•kg-1•min-1) 3) heart rate ±10

176 beats/min of age predicted maximal heart rate (Cooke, 2009). The incremental, ramp treadmill protocol resulted

177 in 100% percent of the participants reaching an RER value of greater than 1.00 and a heart rate value within ±10

178 beats/min of age predicted maximal heart rate, while 88% of the participants presented a plateau in VO 2 despite

179 an increase in speed. It has previously been shown, however, that only a minority of children present a plateau in

180 VO 2 during treadmill exercise (Armstrong & Welsman, 1994) and that children can exercise to exhaustion during

181 incremental exercise without exhibiting a VO 2 plateau (Welsman & Armstrong, 1996). Therefore, as all

182 participants achieved two of the stated criteria and the presence of a VO 2 plateau is not commonly exhibited in

183 children, the term VO 2max was adopted throughout.

184

7
 8

185 The gaseous exchange threshold (GET) was identified using the V-slope method (VCO 2 (ordinate), VO 2 (abscissa)

186 (Beaver, Wasserman & Whipp, 1986). Two regression lines were created based upon the relationship between

187 VCO 2 and VO 2 . The intercept point between the two regression lines was then visually identified, with the VO 2

188 value at the intercept (GET) being extrapolated to the abscissa. To identify the speed at GET, a regression equation

189 was formulated for VO 2 and running velocity, for each individual. The individual’s VO 2 at GET was then inputted

190 into the individual’s respective regression equation to calculate the running velocity at GET. The V-slope method

191 has been shown to be a viable and reliable method for detecting and identifying the gaseous exchange threshold

192 in children (Fawkner, Armstrong, Childs & Welsman, 2002; Unnithan, Timmons, Paton & Rowland, 1995). The

193 GET was assessed by two individual researchers (experienced in the detection of GET), demonstrating 82%

194 agreement (14 out of 17). For the remaining three participants, a third researcher, also experienced in the detection

195 of GET, was approached to verify the GET. Following this test, the treadmill speed corresponding to 95% GET

196 and 60% of the difference between GET and VO 2max (60%∆) were calculated for each individual.

197

198 Assessment of pulmonary oxygen uptake kinetics

199 On the two remaining visits to the laboratory, each participant completed two transitions for each change in

200 workload (unloaded – moderate and moderate – severe), meaning the VO 2 response was averaged from two

201 transitions. Previous research has demonstrated that acceptable confidence intervals for tau when averaging the

202 VO 2 response from two transitions (Marwood et al., 2011; Unnithan et al., 2015). The work-to-work kinetics

203 protocol consisted of 3 min unloaded (standing), 4 min running at a speed equivalent to 95% GET (moderate

204 intensity) and a run to exhaustion at a speed equivalent to 60%∆ (severe intensity). Prior to the test, participants

205 were fitted with a Polar heart rate monitor and face mask, which was connected to an online expired gas analysis

206 system. Participants were also familiarised with the transition to each speed and given sufficient time to practise

207 until they felt comfortable with both transitions (unloaded – moderate and moderate – severe). Participants ran

208 for no longer than 5 sec at the higher intensities during familiarisation. Following familiarisation to the speed

209 transitions, participants rested, allowing their heart rate to return to pre-exercise values, before commencing the

210 test. During the test, verbal encouragement was provided throughout for participants to continue for as long as

211 possible, however no visual feedback relating to exercise duration was given to the participants during the test.

212

8
 9

213 For the unloaded to moderate transition the treadmill was set at the relevant intensity (95% GET; average = 10.4

214 ± 1.0 km/h, range = 8.8 – 12.2 km/h) for each individual, while the participant straddled the treadmill. The

215 participant was then given a 10 second countdown at the end of the unloaded phase, at which point they lowered

216 themselves onto the moving treadmill and began exercising. For the transition from moderate to severe intensity

217 the participant remained running on the treadmill. The time taken for each exercise transition was in all cases < 5

218 seconds, thus having minimal effects on the VO 2 kinetic response as this would be contained within the

219 cardiodynamic phase of the oxygen uptake response to an increase in intensity (Whipp & Rossiter, 2005).

220

221 Mathematical modelling of oxygen uptake kinetics

222 Prior to the modelling of the VO 2 kinetics errant breaths (coughing, swallowing, sighing, etc.) were removed from

223 the raw data set so as not to distort or skew the underlying physiological response. Errant breaths were defined as

224 a breath that was different to the mean of the surrounding four breaths by more than four times the standard

225 deviation of the same surrounding four data points (Jones & Poole, 2005, Marwood et al., 2011). Both data-sets

226 from each stage (unloaded – moderate and moderate – severe) were time aligned and ensemble averaged to

227 enhance the underlying physiological response characteristics for all intensities. Each ensemble average was then

228 linearly interpolated second-by-second prior to the modelling process. Custom written software in Microsoft

229 Excel, using the Solver function, was utilised for all modelling processes.

230

231 Pulmonary oxygen uptake kinetics for the unloaded - moderate and moderate - severe transitions were modelled

232 separately, due to the difference in the characteristics of the kinetic response for each increase in intensity. For

233 both, unloaded – moderate and moderate – to severe transitions, the goal of the modelling process was to isolate

234 the fundamental phase (Whipp & Rossiter, 2005) of VO 2 kinetic response (equation 1). To eliminate the influence

235 of the cardiodynamic phase on the modelling of VO 2 kinetics, the initial 20 seconds from the unloaded to moderate

236 phase and initial 15 seconds of the moderate to severe phase were removed prior to the modelling process. A time

237 delay of 20 seconds is often employed to accommodate the cardiodynamic phase (Whipp & Rossiter, 2005);

238 however, a 15 second time delay was adopted for the moderate to severe transition due to elevated baseline blood

239 flow incurred from the prior moderate intensity (Buchheit, Laursen & Ahmaidi, 2009). Following the

240 cardiodynamic phase, VO 2 kinetics were assumed to develop initially via a single exponential term (fundamental

241 phase), following a delay relative to the start of exercise of the form:

242

9
 10

243 VO 2(t) = VO 2(b) + AVO 2 . (1 – exp - (t-TD)/ τ ) [1]

244

245 Where VO 2(b) is the baseline VO 2 , which was taken as the last 30 seconds of oxygen uptake during 3 minutes

246 unloaded phase or steady state value of moderate for mod – severe transition. AV O2 represents the asymptotic

247 amplitude of the fundamental component of the response; τ is the time constant of the fundamental component

248 and TD is the time delay similar, but not equal to the cardiodynamic-fundamental phase transition time. For the

249 unloaded – moderate transition, the fundamental phase was considered a priori to encapsulate the entire 4-minute

250 transition since exercise was undertaken below the GET. For the moderate – severe transition, the fitting strategy

251 was designed to identify the onset of the “slow component” of the response to exercise, and thus isolate the

252 fundamental component. Starting at 60 s, the fitting window was therefore widened by 1 s until the end of exercise

253 with the time constant and reduced chi-square value of the curve of best fit for each time window plotted against

254 time. The onset of the slow component could then be identified as the coincident point at which a plateau or

255 minima in the value of τ and a minima in chi-square, followed by a progressive increase in these values, could be

256 determined as its value becomes affected by the slow component. The time at which this occurred was used as

257 the optimal fitting window with which to determine the kinetics of the fundamental phase of VO 2 kinetics. The

258 phase III (Whipp & Rossiter, 2005) of oxygen uptake (steady state in the unloaded – moderate transition) was

259 taken as the sum of VO 2(b) and AVO 2 . The amplitude of the slow component during the moderate – severe

260 transition was calculated as VO 2 at exhaustion minus the phase III VO 2 .

261

262 Yo-Yo Intermittent Recovery Test Level 1

263 For the Yo-Yo IR1 test, cones were placed 20 m apart, with a 5 m recovery zone marked out at one end. The Yo-

264 Yo IR1 test requires participants to run 2 x 20 m shuttle runs at increasing speeds, interspersed with 10 seconds

265 of active recovery. The pace of the test was controlled by audio signals emitted from a CD player (Sony CFD-V7,

266 Sony, Tokyo, Japan). For the Yo-Yo IR1 test players were required to run until volitional termination of the test

267 or, when they have twice failed to meet the designated cones in time with the audio signal, at which point they are

268 removed from the test. The test score is the distance covered at the point they withdraw from the test. During the

269 test, players were allowed to consume fluids ad libitum. Current findings support the use of the Yo-Yo IR1 test

270 as a valid measure of physical performance in soccer, particularly within youth populations (Bangsbo et al., 2008;

10
 11

271 Krustrup et al., 2003). All players were familiar with the Yo-Yo IR1, as it is a fitness test regular employed (4

272 times a season) by the respective Academy.

273

274 Monitoring Soccer Match-Play

275 For the assessment of participants’ performance during soccer match-play, three separate, 11 v 11 matches

276 (including the same pitch dimensions, the same composition of players in the same positions and the same

277 coaches) were conducted in the players’ regular training sessions and analysed using 10 Hz GPS and Polar heart

278 rate monitors. As there were only 17 outfield players recruited for the present study, 3 additional outfield players

279 and 2 goalkeepers were used to make up the numbers. These players were not involved within any other aspect of

280 the study and at no point was any data obtained or analysed in respect to these players. Each match was conducted

281 on the same third generation artificial pitch with the same dimensions (90 x 50 m) and at the same time of day.

282 Matches were comprised of 2 x 20 min halves with a 5 min rest interval between halves. The composition of the

283 teams and positions remained the same for all three matches and each participant was assigned their own GPS

284 unit for all matches. Matches were performed on three separate occasions within the same week with a consistent

285 recovery of 48 hrs between matches. Prior to each match, an appropriate and thorough warm-up, as well as a 20

286 minute technical drill was undertaken by all participants. The mean number of satellites during data collection

287 were 8.0 ± 0.5, 8.3 ± 0.4 and 8.2 ± 0.6 for matches 1, 2 and 3 respectively. Furthermore, the mean horizontal

288 dilution of position (HDOP), which is a reflection of the accuracy and quality of the signal (Jennings, Cormack,

289 Coutts, Boyd & Aughey, 2010) were 1.45 ± 0.25, 1.31 ± 0.11 and 1.31 ± 0.08 for matches 1, 2 and 3 respectively.

290

291 Following each match, the GPS data was downloaded and analysed using Catapult Software (Catapult,

292 Melbourne, Australia) and specially designed Microsoft Excel spreadsheets. Following this an average of the 3

293 games was calculated for each player for each variable. Data was recorded for the whole match, each 20 min half

294 and into successive 5 min epochs (e.g. 0 – 5 min, 1 – 6 min, 2 – 7 min, 3 – 8 min, etc), to establish and quantify

295 the peak 5 min epoch (identified as the 5 min epoch with the highest amount of high speed running distance), the

296 subsequent 5 min epoch to the peak 5 min epoch and the mean of the cumulative 5 min epochs throughout the

297 match. These were then averaged over the 3 matches. The peak 5 min epoch provides a surrogate measure of the

298 highest amount of high speed distance that a player can perform within a 5 min period during soccer match-play,

299 whereas the mean of the cumulative 5 min epochs throughout an entire match provides an indirect measure of a

300 player’s ability to sustain their levels of physical performance during match-play. Finally, the subsequent 5 min

11
 12

301 epoch following the peak 5 min epoch provides an estimate of a player’s ability to ‘recover’ and maintain their

302 levels of physical performance, immediately following an exhaustive period of activity. Previous research has

303 used discrete 5 min epochs (0-5 min, 5-10 min, etc.) to identify the peak 5 min interval during soccer match-play

304 (Bradley et al., 2009). The adoption of such criteria, however, may result in missing the true peak 5 min epoch,

305 as this may occur between the pre-determined, discrete 5 min epochs. Consequently, matches were analysed using

306 successive 5 min epochs (0-5 min, 1-6 min, 2-7 min, etc.) for the identification of the most intense 5 min epoch.

307

308 Information recorded included total distance (TD), metres per min (m/min), relative high speed running distance

309 (HSR), relative high speed efforts (HSReff), relative very high speed running distance (VHSR), relative very high

310 speed efforts (VHSReff) and relative sprint distance (S). As noted by Harley et al. (2010) when classifying player

311 motion speed zones, thresholds should be assessed relative to the individuals speed capabilities, with one method

312 being the assessment of maximal velocity obtained during a 20 m sprint. As a result, maximal linear velocity was

313 defined as the maximal velocity obtained during a 20 m straight line sprint from a standing start and obtained

314 from the individual players’ GPS devices, which were then used to record the individual player’s physical

315 performance, using relative thresholds, during soccer match-play. The 20 m straight line sprints were performed

316 as part of the pre-season fitness testing battery (immediately prior to the commencement of this study) employed

317 within the Academy, in which an appropriate warm-up was administered prior to testing. The warm-up consisted

318 of 5 min of dynamic exercise followed by 5 min of dynamic stretches and two trial sprints, however, the

319 environmental factors were not recorded during sprint testing. Research has suggested that distances in the region

320 of 20-30 m are necessary to accurately assess players’ maximal sprint speed in youth populations (Buchheit,

321 Simpson, Peltola & Mendez-Villanueva, 2012), however, previous research from Harley et al. (2010) adopted the

322 same 20 m sprinting protocol, as within the current study, when assessing maximal sprint speed in highly trained

323 youth soccer players. Relative HSR running was regarded as distance covered above 50% of maximal linear

324 velocity, relative VHSR was regarded as any distance covered above 70% of maximal linear velocity and relative

325 Sprint as anything above 90% maximal linear velocity. The same thresholds were used for HSReff and VHSReff

326 and an effort was regarded as any occurrence when such a speed was attained and sustained for greater than 0.2s.

327

328 Statistical Analysis

329 Partial correlations, controlling for maturation using Tanner Stage, between VO 2 kinetics and physical soccer-

330 based measures were performed on the whole sample. To aid interpretation of the results, confidence intervals

12
 13

331 (90%) for correlations were calculated and the magnitude of the correlations were determined using the modified

332 scale by Hopkins (http://www.sportsci.org/resource/stats/, 2000): r < 0.1, trivial; 0.1-0.3, small; 0.3-0.5, moderate;

333 0.5-0.7, large; 0.7-0.9, very large; >0.9, nearly perfect; and 1 perfect). Following this, an intra-group comparison

334 was conducted using independent t-tests, with performance in the maximal YoYo-IR1 being used as the criterion

335 variable (Below Average (BA) n = 9, Above Average (AA) n = 8), as the Yo-Yo IR1 has been shown to be a valid

336 and reliable test with discriminative ability in prospective youth soccer players (Markovic & Mikulic, 2011).

337 Performance variables between groups were compared using Cohen’s d effect sizes (ES) and thresholds (<0.5 =

338 small; 0.5-0.8 = moderate; >0.8 = large; Cohen, 1988) and reported where appropriate. Additionally, where

339 appropriate a qualitative descriptor, used to aid practical inferences, will be assigned to the following quantitative

340 chances of benefit: 25-75% = benefit possible; 75-95% = benefit likely; 95-99% = benefit most likely; >99% =

341 benefit almost certain (Hopkins, 2000). All statistical analysis was performed using SPSS version 21.0 (IBM

342 SPSS statistics for Windows, IBM, Armonk, New York) and Microsoft Excel (Microsoft Excel 2013, Microsoft,

343 Redmond, Washington) with the level of significance (alpha) set at 0.05.

344

345 Results

346 A representative plot of the oxygen uptake response during the work-to-work protocol is shown in figure 1.

347 Moderate inverse correlations were found between the time constant of VO 2 kinetics during the transition to

348 moderate intensity exercise (unloaded – moderate tau) with both field measurements that are associated with

349 physical soccer performance, and relative high intensity activity (HSR & VHSR) during soccer match-play (Table

350 2). However, no relationships were found between measures of VO 2 kinetics, from the moderate – severe

351 transitions and the maximal Yo-Yo IR1 test or measurements obtained during soccer match-play. Furthermore,

352 partial correlations between unloaded – moderate tau and peak 5 min HSR revealed no relationships, partial

353 correlations between unloaded – moderate tau and 5 min HSR match averages revealed significant and moderate

354 inverse correlations (Fig 2a & 2b). Additionally, players’ performance in the maximal Yo-Yo IR1 was found to

355 be significantly related to several measures of physical performance during soccer match-play (Table 3).

356
357 Fig 1 Representative plot of the pulmonary oxygen uptake kinetics during the work-to-work protocol, with the
358 respective unloaded - moderate and moderate - severe time constants (taus)
359

360 *** Insert Figure 1 Here ***

361

362 ***Insert Table 2 Here**

13
 14

363 Fig 2 * P < 0.05, Partial correlations between unloaded – moderate tau and HSR (m) during A) peak 5 min
364 epoch B) match average 5 min epoch
365 *** Insert Figure 2a & 2b Here ***

366

367 ***Insert Table 3 Here***

368

369 Intra-group analysis between AA and BA groups revealed differences (moderate to large Cohen’s d effect sizes)

370 between groups in multiple physical measures during: soccer match-play, field testing and measures of both VO 2

371 kinetics and VO 2max (Table 4a & 4b). With regards to physical measures during soccer-play, only peak 5 min HSR

372 was shown to be ‘unclear’ between groups (according to Hopkins’ (2000) classifications), despite a moderate

373 effect size. With respect to laboratory based measures only VO 2max was shown to be significantly larger in the AA

374 group, however, the tau for pulmonary oxygen uptake kinetics during the transition to moderate intensity exercise

375 was faster in the AA group, revealing a large effect size and ‘likely positive’ inference between groups.

376

377 ***Insert Table 4a & 4b Here***

378

379 Discussion

380 The main finding of the present study was that there was a significant inverse relationship between the time

381 constant (tau) for oxygen uptake kinetics during the transition from unloaded - moderate intensity exercise and

382 the amount of high intensity activities performed during soccer match-play as well as performance in the Yo-Yo

383 IR1. Furthermore, analysis of 5 min epoch match averages revealed significant inverse relationships between

384 unloaded – moderate tau and relative HSR. Finally, results from the intra group analysis revealed that players who

385 performed above average in the Yo-Yo IR1 test demonstrated superior performance in both laboratory tests

386 (unloaded – moderate tau and VO 2max ) and physical performance during soccer match-play (Table 4a & 4b).

387

388 To the authors’ knowledge this is the first study to examine the impact of measures of VO 2 kinetics, along with

389 measures of VO 2max , upon variables of physical performance obtained during soccer match-play, in a group of

390 highly trained youth soccer players. As a result, the present study extends and improves upon existing research

391 via examining the relationship between tau, along with VO 2max , on measures of physical soccer performance in a

392 group of highly trained youth soccer players, during both a standardised test (Yo-Yo IR1) and during soccer

14
 15

393 match-play, using GPS technology. Consequently, this study attempted to adopt a more ecologically valid

394 approach to evaluating physical soccer performance in relation to measures of VO 2 kinetics. This was achieved

395 by assessing players’ physical performance during soccer match-play, rather than the sole use of a surrogate

396 indicator of physical soccer performance (e.g. Yo-Yo IR1). Using this innovative approach, present results have

397 revealed that measures of high intensity activity, obtained during soccer match-play are inversely related to tau,

398 during an unloaded to moderate transition. This supports previous work in both adults and adolescent populations.

399 Firstly, Rampinini et al. (2010) demonstrated that professional adult soccer players demonstrated a faster tau,

400 (from unloaded to moderate transitions) when compared to amateur soccer players (27.2 ± 3.3 vs. 32.3 ± 5.7 s).

401 Secondly, soccer trained male adolescents have been shown to present faster VO 2 kinetics (from unloaded to

402 moderate transitions) when compared to untrained but recreationally active participants (tau: 22.3 ± 7.2 vs. 29.8

403 ± 8.4 s) (Marwood et al., 2010). A finding which is also evident within female adolescent soccer players, when

404 compared to untrained but recreationally active participants (tau: 26.3 ± 6.9 vs. 35.1 ± 11.5 s) (Unnithan et al.,

405 2015). Current findings, however, expand upon previous research by highlighting the impact of superior VO 2

406 kinetics upon key measures of physical soccer performance (e.g. HSR activity), during match-play and within a

407 group of highly trained youth soccer players. Consequently, as levels of high intensity activity have been shown

408 to be a distinguishing factor of superior soccer performance (Mohr, Krustrup & Bangsbo 2003), an improved

409 ability to accommodate the necessary energy demands, during soccer match-play, via superior VO 2 kinetics and

410 utilisation of aerobic pathways may result in a greater accumulation, frequency and maintenance of high intensity

411 activity.

412

413 Within the assessment of players’ physical performance during soccer match-play, a secondary aim was to assess

414 the impact of VO 2 kinetics on players’ physical performance during the peak 5 min epoch, post-peak 5 min epoch

415 and the match average 5 min epochs. Results demonstrate that while measures of VO 2 kinetics, during an unloaded

416 – moderate and moderate – severe transition, have little influence upon the amount of high intensity activity

417 performed over a short period of time (peak 5 min) or in the subsequent 5 min period (post-peak 5 min), measures

418 of unloaded-moderate tau were significantly related to 5 min epoch match averages for measures of HSR, with a

419 faster unloaded – moderate tau equating to a greater 5 min epoch match average for HSR. Thus, there is evidence

420 to suggest that measures of VO 2 kinetics, particularly unloaded - moderate tau, may have a significant impact

421 upon the maintenance and the ability to sustain levels of high intensity activity over a prolonged period of time

15
 16

422 (20 - 40 min), during competitive soccer match-play. This is in line with Dupont et al. (2005), who reported a

423 significant relationship between the phase II tau, during a low to moderate transition, and the percentage

424 decrement in speed during 15 repeated 40 m maximal sprints, with 25 s of active recovery. Together, these results

425 demonstrate that when numerous bouts of repeated high intensity activity are required, superior VO 2 kinetics are

426 potentially advantageous to performance via a greater tolerance to mechanisms of fatigue (Buchheit, Abiss, Peiffer

427 & Laursen, 2012; Dupont et al., 2005). However, the exercise prescribed in the study by Dupont et al. (2005)

428 greatly exceeds the work that would be experienced during an intense 5 min period of competitive soccer match-

429 play. Rather, the volume of work prescribed by Dupont et al. (2005) is indicative of the amount of work performed

430 during a prolonged period of competitive soccer match-play. Nevertheless, current findings suggest that, while

431 measures of VO 2 kinetics, during an unloaded to moderate transition, may not have an impact upon the high

432 intensity activity over a short period of time (5 min) during competitive soccer match-play, they do have an

433 increasing influence over a prolonged period of time. This suggests that superior VO 2 kinetics preserve and help

434 to maintain intermittent high intensity activity during sustained periods of soccer match-play, a finding which

435 may be linked to the enhanced ability to activate the more sustainable aerobic energy sources.

436

437 While the present results provide some evidence for the relationship between physical measures associated with

438 superior soccer performance and VO 2 kinetics, the complex physiology involved during high-intensity

439 intermittent exercise should be acknowledged. As a result research in this area has produced equivocal results,

440 with some studies finding little or no relationship between VO 2 kinetics and intermittent exercise performance

441 (Buchheit, Hader & Mendez-Villanueva, 2012; Christensen et al., 2011; Wells et al., 2012) and other studies

442 demonstrating significant associations between superior VO 2 kinetics and intermittent exercise capacity (Dupont

443 et al., 2005; Rampinini et al., 2009; Rampinini et al., 2010). Performance during high-intensity intermittent

444 exercise, like that which is experienced during soccer match-play, is a product of complex interactions between

445 metabolic, cardiorespiratory and cardiovascular issues. Indeed, factors such as transportation of ions, metabolite

446 accumulation, muscle excitability and muscle oxygenation (Girard, Mendez-Villanueva & Bishop, 2011) among

447 numerous other factors will impact upon the presence of fatigue, and therefore the level of physical performance

448 during soccer match-play (Bangsbo et al., 2006).

449

16
 17

450 Consequently, one variable (e.g. tau or VO 2max ) will not, on its own, predict players’ intermittent exercise capacity,

451 as is evidenced by the correlational statistics provided within the present study. Furthermore, the finding that the

452 AA Yo-Yo group performed significantly more HSR and VHSR than the BA group, despite there being no

453 significant difference in measures of tau, supports the fact that performance during high-intensity intermittent

454 exercise is multifaceted. Nevertheless, the current results provide an indication to the contribution of the energy

455 systems during youth soccer match-play, demonstrating that a faster tau, during an unloaded to moderate

456 transition, was associated to the accumulation, frequency and maintenance of high intensity activity during soccer

457 match-play. Additionally, current results reinforce the adoption of the Yo-Yo IR1 test as an appropriate test for

458 assessing sport-specific fitness levels in highly trained youth soccer players.

459

460 Therefore, methods of training which are capable of speeding up tau, should improve the accumulation, frequency

461 and maintenance of high intensity activity during soccer match-play, in highly trained youth soccer players.

462 Indeed, high intensity conditioning drills or the use of game-based training drills (e.g. small sided games), which

463 are designed to provide sustained periods (3 - 4min) at an intensity >85% HR max , at a work to rest ratio of 2:1,

464 may be incorporated as a stimulus to enhance players’ aerobic capacity and subsequently improve measures of

465 VO 2 kinetics. Similarly, the use of supra-maximal repeated sprint drills or drills which require a player to work

466 above their maximal aerobic speed for shorter intervals may be also utilised as a conditioning method for

467 improving players’ VO 2 kinetics (Buchheit & Laursen, 2013).

468

469 There were no statistically significant relationships between measures of VO 2 kinetics (tau) during the moderate

470 to severe transitions and any of measures associated with physical soccer performance. Current findings are

471 supported by those of Nyberg et al. (2016), who reported no change in measures of VO 2 kinetics during a moderate

472 (10 km/h) to severe (16 km/h) transition in highly trained adult soccer players, following a period of additional

473 speed endurance training. This was despite significant improvements in Yo-Yo IR1 performance (11.6 ± 6.4%)

474 and pulmonary VO 2 kinetics during an unloaded to moderate transition (11.4 ± 16.5%) (Nyberg et al., 2016).

475

476 In addition, the current results show a large standard deviation is apparent for measures of tau during the moderate

477 - severe transition, these large variations in tau may be a result of the intensity prescribed for severe exercise

478 (60%Δ). The prescribed intensity of 60%Δ, despite being a relative intensity, is likely to be in the proximity of

479 critical speed (Burnley & Jones, 2007; Ozyener, Rossiter, Ward & Whipp, 2001), resulting in some individuals

17
 18

480 being above their critical speed and some being below. As a result, this may have elicited different physiological

481 responses across the current sample. The times recorded, however, for time to exhaustion during the severe

482 intensity (551.6 ± 167.5 s) provide support that all players were performing at an intensity above their critical

483 speed. Nevertheless, the prescribed intensity may have resulted in some individuals not being able to maintain

484 their performance for a satisfactory amount of time, thus affecting the ‘fit’ of the VO 2 response. As noted by

485 Ozyener et al. (2001) the VO 2 kinetics during dynamic muscular exercise are clearly influenced by the exercise

486 intensity. Research also suggests there is a reduced VO 2 amplitude in the slow component when running compared

487 to cycling exercise (Carter et al., 2000; Jones & McConnell, 1999). Furthermore, the breath-to-breath noise has

488 been shown to be large in child populations (Potter, Childs, Houghton & Armstrong, 1998) while the amplitude

489 of the VO 2 slow component has been shown to be small in child populations (Williams et al., 2001). This can

490 result in low signal to noise properties within the VO 2 profiles and impact upon the fitting of the moderate - severe

491 transition, making it difficult to reliably identify the non-steady state ‘phase’ of the VO 2 response during this

492 transition (Armstrong & Barker, 2009). To date, there is limited knowledge as to whether or not trained individuals

493 actually present faster VO 2 kinetics during work-work transitions and, therefore, the applicability of VO 2 kinetics

494 profiles obtained during a moderate to severe transition and their relationship to physical soccer performance

495 requires further research. Although, initial results from the current study and that of Nyberg et al. (2016) would

496 seem to indicate that the VO 2 kinetics profile obtained from a moderate to severe transition is unrelated and not

497 influenced by the high-intensity intermittent exercise that is experienced during soccer match-play.

498

499 Previous research that has examined the relationship between VO 2 kinetics and physical measures associated with

500 soccer match-play has failed to provide physical data from performance during competitive soccer match-play.

501 The addition of such information will improve the ecological validity and application of measures obtained from

502 both laboratory and field tests to team sports performance. It is understood, however, that reasonable levels of

503 game to game variation exist within soccer match-play (Gregson et al., 2010). Indeed, Gregson et al. (2010)

504 reported high levels of variance for HSR (CV = 16.2 ± 6.2%) during elite adult competitive football when

505 analysing players’ physical performance across two seasons. Information regarding the variance in physical

506 performance, for youth soccer players is lacking, however, analysis of the variance across the three matches

507 employed within the current study revealed typical errors of 4.3 m/min for TD, 316 m & 24 efforts in the HSR

508 threshold and 75 m & 5 efforts in the VHSR threshold. While the values presented are substantial, when applied

18
 19

509 in relation to the data reported for the AA and BA groups, within the current study, the difference between groups

510 is in excess of the typical errors presented.

511

512 In previous research, the resultant impact of the opposing team (Gabbett, 2013), phase of the season (Gregson et

513 al. 2010; Kempton, Sullivan, Bilsborough, Cordy & Coutts, 2015), weather conditions, substitutions, context of

514 the match (win/lose margin) and current form (Black & Gabbett, 2014) are all likely to have an influence on

515 players’ physical performance and need to be considered when analysing players’ physical performance during

516 soccer match-play. Nevertheless, the current procedures controlled for the composition of players, positions and

517 teams as well as the playing surface and playing time, in an attempt to reduce the variance from game-to-game.

518

519 Finally, the potential impact of growth and maturation should not be ignored when assessing youth soccer players.

520 In the present study, however, the effects of growth and maturation upon the fitness components measured was

521 equivocal. Indeed, no relationships between maturation status, measured via Tanner Stage and Maturity Offset,

522 were found between measures of VO 2 kinetics and measures associated with physical performance during soccer

523 match-play. Although, the limitations of the adopted methods used to assess growth and maturation do need to be

524 acknowledged. Firstly, the Tanner stage method of maturation does not provide information on the point at which

525 a player entered a particular stage of maturation or how long an individual has been in a particular stage of

526 maturation (Malina et al., 2012). Secondly, the accuracy of the predictive equation used to determine the maturity

527 off-set has been shown to decrease the further the individual is away from the period of peak height velocity

528 (Malina, Cumming, Morano, Barron & Miller, 2005). Nevertheless, the impact of growth and maturation within

529 the current sample, particularly during the intra-group analysis, was negligible. Furthermore, partial correlations

530 were employed throughout to control for any possible influence of growth and maturation. Despite these

531 limitations the current study extends and improves upon the existing research by examining distinct measures of

532 cardio-respiratory fitness in relation to measures of physical performance obtained during soccer match-play in

533 elite youth soccer players.

534

535 Conclusion

536 The results from this study demonstrate that players with a superior ability to activate aerobic energy pathways at

537 the onset of exercise, evidenced by a faster tau during transitions from low to moderate intensity exercise, tend to

538 perform better in Yo-Yo IR1 and perform a greater amount of relative HSR during soccer match-play. Hence, the

19
 20

539 accumulation, maintenance and frequency of high intensity activity during competitive soccer match-play, in

540 highly trained youth soccer players, are associated with a faster tau during an unloaded to moderate exercise

541 transition but unrelated to tau during a moderate – severe exercise transition. This study demonstrates that

542 measures obtained from VO 2 kinetics testing protocols may be of benefit when assessing youth players’ physical

543 performance during soccer match-play and could therefore be used as a measure to distinguish those of superior

544 physical performance. Based on these findings, the physiological capacity to transition rapidly between

545 workloads, particularly from low – moderate intensities, is a potential determinant of superior physical

546 performance (specifically high intensity activities), during soccer match-play, in highly trained youth soccer

547 players.

548

549 References

550 1. Armstrong N, Barker A (2009) Oxygen uptake kinetics in children and adolescents: A review. Ped Exer

551 Sci 21(2):130-147.

552 2. Armstrong N, Welsman J (1994) Assessment and interpretation of aerobic fitness in children and

553 adolescents. Exer Sports Sci Rev 22(1):435-76.

554 3. Astorino T, Robergs R, Ghiasvand F, Marks D, Burns S (2000) Incidence of the oxygen plateau at

555 VO 2max during exercise testing to volitional exhaustion. J Exer Physiol Online 3(4):1-12.

556 4. Bangsbo J, Iaia M, Krustrup P (2008) The Yo-Yo intermittent recovery test: A useful tool for evaluation

557 of physical performance in intermittent sports. Sports Med 38(1):37-51. doi:10.2165/00007256-

558 200838010-00004.

559 5. Bangsbo J, Mohr M, Krustrup P (2006) Physical and metabolic demands of training and match-play in

560 the elite football player. J Sports Sci 24(7):665-74. doi:.org/10.1080/02640410500482529.

561 6. Beaver WL, Wasserman K, Whipp BJ (1986) A new method for detecting anaerobic threshold by gas

562 exchange. J Appl Physiol 60(6):2081-2027.

563 7. Black, G., & Gabbett, T. (2014). Match intensity and pacing strategies in rugby league: an

564 examination of whole-game and interchanged players, and winning and losing teams. Journal of

565 Strength & Conditioning Research, 28(6), 1507-1516. doi: 10.1519/JSC.0b013e3182a4a225.

566 8. Bradley P, Sheldon W, Wooster B, Olsen P, Boanas P, Krustrup P (2009) High-intensity running in

567 English FA Premier League soccer matches. J Sports Sci 27(2):159-68.

568 doi:10.1080/02640410802512775.

20
 21

569 9. Buchheit M, Abiss CR, Peiffer JJ, Laursen, PB (2012) Performance and physiological responses during

570 a sprint interval training session: relationships with muscle oxygenation and pulmonary oxygen uptake

571 kinetics. Eur J Appl Phys 112(2):767-79. doi:10.1007/s00421-011-2021-1.

572 10. Buchheit M, Hader K, Mendez-Villanueva A (2012) Tolerance to high-intensity intermittent running

573 exercise: do oxygen uptake kinetics really matter? Frontiers in Physiology 3(406):1-13.

574 doi:10.3389/fphys.2012.00406.

575 11. Buchheit M, Laursen P (2013) High-intensity interval training, solutions to the programming puzzle:

576 Part 2; anaerobic energy, neuromuscular load and practical applications. Sports Med 43(5):313-338.

577 doi:10.1007/s40279-013-0029-x.

578 12. Buchheit M, Laursen PB, Ahmaidi, S (2009) Effect of prior exercise on pulmonary O 2 uptake and

579 estimated muscle capillary blood flow kinetics during moderate intensity field running in men. J Appl

580 Physiol 107(2):460-70. doi:10.1152/japplphysiol.91625.2008.

581 13. Buchheit M, Simpson B, Peltola E, Mendez-Villanueva A (2012) Assessing maximal sprinting speed in

582 highly trained young soccer players. Int J Sports Physio Perf, 7:76-78.

583 14. Burnley M, Jones AM (2007) Oxygen uptake kinetics as a determinant of sports performance. Eur J

584 Sport Sci 7(2):63-79. doi:10.1080/17461390701456148.

585 15. Carter H, Jones AM, Barstow TJ, Burnley M, Williams CA, Doust JH (2000) Oxygen uptake kinetics

586 in treadmill running and cycle ergometry: a comparison. J Appl Physiol. 89(3):899-907.

587 16. Christensen P, Krustrup P, Gunnarsson T, Kiilerich K, Nybo L, Bangsbo J (2011) VO 2 kinetics and

588 performance in soccer players after intense training and inactivity. Med Sci Sports Exerc. 43(9):1716-

589 1724. doi: 10.1249/MSS.0b013e318211c01a.

590 17. Cohen J (1988) Statistical power analysis for the behavioural sciences 2nd ed. Hillsdale (NJ): Lawrence

591 Erlbaum Associates; p. 567.

592 18. Cooke C (2009) Maximal Oxygen Uptake, Economy and Efficiency. In: Eston R & Reilly T, editors.

593 Kinanthropometry and Exercise Physiology Laboratory Manual: Tests, Procedures and Data (3rd

594 Edition). London: Routledge, pp. 174-212.

595 19. Cunha G, Lorenzi T, Sapata K, Lopes AL, Gaya AC, Oliveira A. (2011). Effect of biological maturation

596 on maximal oxygen uptake and ventilatory thresholds in soccer players: An allometric approach. J

597 Sports Sci. 29(10):1029-1039. doi: 10.1080/02640414.2011.570775.

21
 22

598 20. Deprez D, Buchheit M, Fransen J et al. (2015). A longitudinal study investigating the stability of

599 anthropometry and soccer-specific endurance in pubertal high-level youth soccer players. J Sports Sci

600 & Med. 14(2): 418-426.

601 21. Deprez D, Coutts A, Lenoir M et al. (2014). Reliability and validity of the Yo-Yo intermittent recovery

602 test level 1 in young soccer players. J Sports Sci. 32(10):903-910. doi: 10.1080/02640414.2013.876088.

603 22. Dupont G, Millet G, Guinhouya C, Berthoin S (2005) Relationship between oxygen uptake kinetics and

604 performance in repeated running sprints. Eur J Appl Phys 95(1):27-34. Doi:10.1007/s00421-005-1382-

605 8.

606 23. Durnin, J, Womersley, J. (1974). Body fat assessed from total body density and its estimation from

607 skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. British Journal of

608 Nutrition, 32: 77-97. doi:10.1079/BJN19740060.

609 24. Fawkner S, Armstrong N, Childs DJ, Welsman JR (2002) Reliability of the visually identified ventilatory

610 threshold and V-slope method in children. Pediatric Exercise Science 14(2):181-192.

611 25. Gabbett, T. (2013). Influence of the opposing team on the physical demands of elite rugby league

612 match play. Journal of Strength & Conditioning Research, 27(6), 1629-1635.

613 doi:10.1519/JSC.0b013e318274f30e.

614 26. Gellish RL, Goslin BR, Olson RE, McDonald A, Russi G, Moudgil VK (2007) Longitudinal Modeling

615 of the relationship between age and maximal heart rate. Med Sci Sports Exerc 39(5):822-29.

616 doi:10.1097/mss.0b013e31803349c6.

617 27. Girard O, Mendez-Villanueva A, Bishop D (2011) Repeated sprint ability – Part 1: Factors contributing

618 to fatigue. Sports Med 41(8):673-694. doi:10.2165/11590550-000000000-00000.

619 28. Glaister, M. (2005). Multiple Sprint Work: Physiological responses, mechanisms of fatigue and the

620 influence of aerobic fitness. Sports Medicine. 35: 757-777. doi:10.2165/00007256-200535090-00003.

621 29. Gregson W, Drust B, Atkinson G, Salvo V (2010) Match-to-match variability of high speed activities in

622 premier league soccer. Int J Sports Med 31(4):237-242. doi:10.1055/s-0030-1247546.

623 30. Harley J, Barnes C, Portas M, Lovell R, Barrett S, Paul D, Weston M (2010) Motion analysis of match-

624 play in elite U12 to U16 age-group soccer players. J Sports Sci. 28(13):1391-1397.

625 doi:10.1080/02640414.2010.510142.

626 31. Hopkins WG (2000) A new view of statistics. Internet Society for Sport Science.

627 http://www.sportsci.org/resource/stats/.

22
 23

628 32. Iaia, FM, Rampinini E, Bangsbo J (2009) High-intensity training in football. Int J Sports Physiol Perf

629 4(3):291-306.

630 33. Jennings D, Cormack S, Coutts A, Boyd L, Aughey J (2010) The validity and reliability of GPS units

631 for measuring distance in team sport specific running patterns. Int J Sports Physiol Perf, 5(3):328-341.

632 http://hdl.handle.net/10453/16884.

633 34. Jones AM, Doust JH (1996) A 1% treadmill grade most accurately reflects the energetic cost of outdoor

634 running. Journal of Sports Sciences 14(4):321-27.

635 35. Jones AM, McConnell AM (1999) Effect of exercise modality on oxygen uptake kinetics during heavy

636 exercise. Eur J Appl Physiol 80(3):213-219.

637 36. Jones AM, Poole DC (2005) Introduction to Oxygen Uptake Kinetics. In: Jones AM, Poole DC, editors.

638 Oxygen Uptake Kinetics in Sport, Exercise & Medicine. London: Routledge, p. 3-35.

639 37. Kempton, T., Sullivan, C., Bilsborough, J., Cordy, J., & Coutts, A. (2015). Match-to-match variation

640 in physical activity and technical skill measures in professional Australian Football. Journal of Science

641 & Medicine in Sport, 18(1), 109-113. doi:10.1016/j.jsams.2013.12.006.

642 38. Krustrup P, Mohr M, Amstrup T et al. (2003) The Yo-Yo intermittent recovery test: Physiological

643 response, reliability, and validity. Med Sci in Sports Exerc 35(4):697-705.

644 doi:10.1249/01.MSS.0000058441.94520.32.

645 39. Le Gall F, Carling C, Williams M, Reilly T (2010) Anthropometric and fitness characteristics of

646 international, professional and amateur male graduate soccer players from an elite youth academy. J Sci

647 & Med Sport. 13:90-95. doi:10.1016/j.jsams.2008.07.004.

648 40. Malina R, Coelho e Silva M, Figueiredo J et al. (2012) Interrelationships among invasive and non-

649 invasive indicators of biological maturation in adolescent male soccer players. J Sports Sci 30(15):1705-

650 1717. doi:10.1080/02640414.2011.639382.

651 41. Malina R, Cumming S, Morano P, Barron M, Miller S (2005) Maturity status of youth football players:

652 a noninvasive estimate. Med Sci Sports Exerc 37(6):1044-1052.

653 doi:10.1249/01.mss.0000171622.45134.cb.

654 42. Marwood S, Jack S, Patel M, Walker P, Bowtell J, Calverley P (2011) No effect of glutamine ingestion

655 on indices of oxidative metabolism. Respiratory Physiology and Neurobiology 177(1), 41-6.

656 doi:10.1016/j.resp.2011.03.010.

23
 24

657 43. Marwood S, Roche D, Rowland T, Garrard M, Unnithan VB (2010) Faster pulmonary oxygen uptake

658 kinetics in trained versus untrained male adolescents. Med Sci Sports Exerc 42(1):127-34.

659 doi:10.1249/MSS.0b013e3181af20d0.

660 44. Markovic G, Mikulic P (2011) Discriminative ability of the Yo-Yo intermittent recovery test (level 1)

661 in prospective young soccer players. J Strength Cond Res 25(10):2931-2934.

662 doi:10.1519/JSC.0b013e318207ed8c.

663 45. Mirwald R, Baxter-Jones A, Bailey D, Beunen G (2002) An assessment of maturity from anthropometric

664 measurements. Med Sci Sports Exerc 34(4):689-694. doi:10.1097/00005768-200204000-00020.

665 46. Mohr M, Krustrup P, Bangsbo, J (2003) Match performance of high standard soccer players with special

666 reference to development of fatigue. J Sports Sci 21(7):519-528. doi:10.1080/0264041031000071182.

667 47. Nyberg, M, Fiorenza, M, Lund, A et al. (2016) Adaptations to speed endurance training in highly trained

668 soccer players. Med Sci Sports Exerc. Published ahead of print.

669 48. Ozyener F, Rossiter HB, Ward SA, Whipp BJ (2001) Influence of exercise intensity on the on- and off-

670 transient kinetics of pulmonary oxygen uptake in humans. J Physiol 53(3):891-902.

671 49. Potter C, Childs D, Houghton W, Armstrong, N (1999). Breath-to-breath “noise” in the ventilatory and

672 gas exchange responses of children to exercise. Eur J Appl Phys 80(2):118-124.

673 50. Rampinini E, Sassi A, Azzalin A et al. (2010) Physiological determinants of Yo-Yo intermittent

674 recovery tests in male soccer players. Eur J Appl Phys 108(2):401-9. doi:10.1007/s00421-009-1221-4.

675 51. Rampinini E, Sassi A, Morelli A, Mazzoni S, Fanchini M, Coutts (2009) Repeated-sprint ability in

676 professional and amateur soccer players. Appl Physiol Nutr Metab 34(6):1048-1054. doi:10.1139/H09-

677 111.

678 52. Rebelo A, Brito J, Seabra A, Oliveira J, Krustrup P (2014) Physical match performance of youth football

679 players in relation to physical capacity. Eur J Sport Sci 14(S1):S148-S156.

680 doi:10.1080/17461391.2012.664171.

681 53. Stroyer J, Hansen L, Klausen K (2004) Physiological profile and activity pattern of young soccer players

682 during match play. Med Sci Sports Exerc 36(1):168-174. doi:10.1249/01.MSS.0000106187.05259.96.

683 54. Tanner J. Growth at adolescence (2nd ed.). Oxford: Blackwell Scientific, 1962.

684 55. Unnithan VB, Roche D, Garrard M, Holloway K, Marwood S (2015) Oxygen uptake kinetics in trained

685 adolescent females. Eur J Appl Phys 115(1):213-20.doi:10.1007/s00421-014-3005-8.

24
 25

686 56. Unnithan VB, Timmons JA, Paton JY, Rowland TW (1995) Physiologic correlates to running

687 performance in pre-pubertal distance runners. Int J Sports Med 16(8):528-533.

688 57. Wells C, Edwards A, Winter E, Fysh M, Drust B (2012) Sport-specific fitness testing differentiates

689 professional from amateur players where VO 2max and VO 2 kinetics do not. J Sports Med Phys Fitness

690 52:245-254.

691 58. Welsman JR, Armstrong, NA (1996) The measurement and interpretation of aerobic fitness in children:

692 current issues. Journal of the Royal Society of Medicine 89(5):281P-285P.

693 59. Whipp BJ, Rossiter HB (2005) The Kinetics of Oxygen Uptake: Physiological Inferences from the

694 Parameters. In: Jones AM, Poole DC, editors. Oxygen Uptake Kinetics in Sport, Exercise & Medicine.

695 London: Routledge, p. 62-94.

696 60. Williams CA, Carter H, Jones AM, Doust JH (2001) Oxygen uptake kinetics during treadmill running

697 in boys and men. J Appl Physiol 90(5):1700-6.

25