Articles in PresS. J Appl Physiol (March 21, 2013). doi:10.1152/japplphysiol.00575.

2012

1 High-Intensity Interval Training Speeds the Adjustment of Pulmonary O2

2 Uptake, but not Muscle Deoxygenation, During Moderate-Intensity Exercise

3 Transitions Initiated from Low and Elevated Baseline Metabolic Rates

5 Alexandra M. Williams1,2, Donald H. Paterson1,2, and John M. Kowalchuk1,2,3

7 1 Canadian Centre for Activity and Aging and School of Kinesiology, The University
8 of Western Ontario, London, ON, Canada

9 2 School of Kinesiology, The University of Western Ontario, London, ON, Canada

10 3 Department of Physiology and Pharmacology, The University of Western

11 Ontario, London, ON, Canada

12

13 Running head: HIT Speeds VO2p Adjustments from Elevated Metabolic Rates

14

15 Corresponding author: Dr. J.M. Kowalchuk

16 School of Kinesiology
17 University of Western Ontario
18 London, Ontario, Canada
19 N6A 3K7

20 e-mail: jkowalch@uwo.ca

21

22 Author Contributions: A.M.W., D.H.P and J.M.K. were all involved in the
23 conception and design of the project. A.M.W. ran and supervised all training
24 sessions, performed all data collection and analysis. A.M.W. and J.M.K.
25 interpreted the data. A.M.W. wrote the manuscript, with input and revisions
26 from J.M.K. and D.H.P. All training and experiments were performed in the
27 laboratories of J.M.K and D.H.P.

28

Copyright © 2013 by the American Physiological Society.

 2

29 Abstract

30 During step-transitions in work rate (WR) within the moderate-intensity (MOD)

 
31 exercise domain, pulmonary O2 uptake ( V O 2p ) kinetics are slowed and V O 2p gain (Δ

 O 2p
V
32 /ΔWR) is greater when exercise is initiated from an elevated metabolic rate.


33 High-intensity interval training (HIT) has been shown to speed V O 2p kinetics when

34 step-transitions to MOD exercise are initiated from light-intensity baseline metabolic

35 rates. The effects of HIT on step-transitions initiated from elevated metabolic rates

36 have not been established. Therefore, this study investigated the effects of HIT on

37  O 2p kinetics during transitions from low and elevated metabolic rates, within the
V

38 MOD domain. Eight young, untrained men completed 12 sessions of HIT (spanning 4

39 weeks). HIT consisted of 8-12 1-min intervals, cycling at a WR corresponding to

40 110% of pre-training WRmax. Pre-, mid- and post-training, subjects completed a ramp-

41  O 2max , WR and estimated lactate threshold ( θ̂ L ).

incremental test to determine V max

42 Participants additionally completed double-step constant-load tests, consisting of step

43 transitions from 20W→Δ45% θ̂ L (lower step; LS) and Δ45→90% θ̂ L (upper step;

44  O 2max (P<0.05) and WRmax (P<0.01), and τ V O 2p of

US). HIT led to increases in V

45 both lower and upper MOD step transitions were reduced by ~40% (LS: 24s→ 15s;

46 US: 45s→ 25s) (P<0.01). However, the time course of adjustment of local muscle

47 deoxygenation was unchanged in the LS and US. These results suggest that speeding


48 of V O 2p kinetics in both the LS and US may be due, in part, to an improved matching

49 of muscle O2 utilization-to-microvascular O2 delivery within the working muscle

50 following 12 sessions of HIT, although muscle metabolic adaptations cannot be

 3

51 discounted.

52

53 Key Words: O2 uptake kinetics, near-infrared spectroscopy (NIRS), O2 deficit,

54 exercise performance

55
 4

56 Introduction

57 Coincident with a step increase in exercise intensity there is an immediate

58 increase in ATP turnover in the working muscle (52). Pulmonary O2 uptake ( V O 2p ),

59 however, does not increase instantaneously, but rather, rises with an exponential time

60 course towards a new steady-state level, after allowing for the transport delay

61 between the active muscle and the pulmonary circulation (4, 25, 35). The kinetics of

62 the fundamental phase II V O 2p response predominantly reflects the adjustment of

63 mitochondrial O2 utilization in the working muscle (4, 25, 35). Additional energy

64 requirements not met by oxidative phosphorylation are met through substrate-level

65 phosphorylation with consequent breakdown of phosphocreatine (PCr) and glycogen

66 (25, 33, 52).

67 When exercise transitions to work rates (WR) within the moderate- (MOD) (8,

68 9, 32, 40, 55), or the heavy- or severe-intensity domains (17, 32, 60, 61), are initiated

69 from an elevated compared to lower baseline WR and thus metabolic rate, the time

70 course of adjustment of V O 2p is slower than when starting from a low or resting

71 metabolic rate. Additionally, the fundamental V O 2p gain (Δ V O 2p /ΔWR), a measure

72 of the O2 cost of the change in work rate (reflecting exercise efficiency), is greater for

73 exercise initiated from a higher compared to lower metabolic rate (8, 9, 40, 55, 60,

74 61). Although the specific mechanisms governing the observed slowed V O 2p kinetics

75 and lower efficiency for exercise transitions initiated from an elevated baseline

76 metabolic rate with the MOD domain have not been established, the differing

77 responses may reflect intrinsic properties of newly recruited muscle fibres (i.e.,

78 having a slower activation and lower efficiency) (9); a less favourable cellular
 5

79 energetic state of the active fibres (lower cellular PO2, [PCr] and [ATP]; higher

80 [ADP] and [Pi]; less negative ∆GATP) (42); a slower adjustment of central (cardiac

81 output) and peripheral conduit artery bulk blood flow and O2 delivery, and/or a lower

82 steady-state increase in blood flow relative to muscle O2 utilization (32, 40); or a

83 combination of these factors.

84 The performance of low-volume, high-intensity interval training (HIT) (or

85 low-volume, high-intensity sprint interval training, SIT) promotes rapid adaptations in

86 maximal V O 2p , work capacity, muscle oxidative capacity and vascular function (3,

87 12, 22, 30, 36, 41, 49), with consequent improvements in exercise performance (3, 13,

88 22, 41). For example, Burgomaster et al. (11, 13) observed increases in resting

89 muscle glycogen, mitochondrial citrate synthase (CS) and pyruvate dehydrogenase

90 (PDH) activity, and improved time trial (TT) performance following 6 sessions (2

91 weeks) of SIT. Gibala et al. (23) reported immediate increases in markers of

92 mitochondrial biogenesis following a single session of SIT (total of 2 min; <80kJ total

93 work). Recently, McKay et al. (41) reported that the adjustment of V O 2p during

94 single-step transitions into the upper part of the MOD domain (20 W → 90% lactate

95 threshold ( θ̂ L )) became faster after only 2 sessions of HIT (τ V O 2p reduced by

96 ~20%), with an even faster adjustment of V O 2p seen after 8 training sessions (τ V O 2p


97 reduced by ~40%). Additionally, the speeding of V O 2p kinetics (and presumably

98 muscle O2 utilization kinetics) with HIT occurred in the absence of any changes in

99 muscle deoxygenation kinetics (41), implying that faster adjustments of local muscle

100 (microvascular) blood flow and O2 delivery accompanied the faster rate of adjustment

101 of muscle O2 utilization. These findings clearly illustrate the effectiveness of HIT in
 6

102 producing rapid improvements to skeletal muscle oxidative potential and exercise

103 performance, and faster adjustments in oxidative ATP production.

104 Although a faster adjustment of V O 2p kinetics during exercise initiated from a

105 low baseline metabolic rate is expected following HIT (7, 41, 47) the effect of HIT

106 (or more traditional types of endurance training) on exercise transitions initiated from

107 an elevated baseline metabolic rate has not been examined. Additionally, previous


108 reports from our laboratory reported greater acute speeding of V O 2p kinetics within

109 the MOD domain following a priming bout of heavy-intensity exercise for those


110 individuals presenting with slower V O 2p kinetics in an “unprimed” condition (27-29,

111 53). Thus, we questioned whether the rapid HIT-induced adaptations in muscle

112 oxidative capacity and vascular function seen in prior reports (12, 22, 30, 36, 49)

113 would result in a greater speeding of V O 2p kinetics, without affecting muscle

114 deoxygenation kinetics (i.e., reflecting the dynamic relationship between local

115 microvascular blood flow and muscle O2 utilization), during exercise transitions in

116 the US compared to the LS.

117 The primary goal of this study was to investigate the effects of HIT on V O 2p

118 and muscle deoxygenation kinetics during transitions from low and elevated

119 metabolic rates, within the MOD domain. Single-step transitions from the low

120 metabolic rate to the higher WR and time-to-fatigue endurance trials were

121 additionally examined, in order to compare (and confirm) both to prior training

122 studies and to the training-induced changes in the LS and US of the current study.

123 The following hypotheses were tested: 1) HIT would speed V O 2p kinetics in both the

124 LS and US without affecting muscle deoxygenation ([HHb]) kinetics, 2) HIT-induced

 7

125 speeding of V O 2p kinetics would be greater in the US compared to LS, as

126 consequence of attenuated V O 2p kinetics in the US in the untrained state, and 3) a

127 transient overshoot in the ratio of muscle deoxygenation-to- V O 2p would be present

128 before but not after HIT as a consequence of an improved dynamic relationship

129 between the adjustment of microvascular blood flow and muscle O2 utilization.

130

131 Methods

132 Subjects

133 Eight young, healthy adult men were recruited to complete the high-intensity

134 interval training program (HIT; 27 ± 6 y (mean ± SD), 82 ± 5 kg). An additional 5

135 young, healthy men served as control subjects (CON; 23 ± 3 y, 79 ± 9 kg) and

136 completed all aspects of the experimental protocol except for the high-intensity

137 interval training. All participants were untrained, recreationally active (other light-to-

138 moderate intensity activities, up to 2-3 times per week), and were asked to continue

139 their regular daily activities for the duration of the study. All participants reported

140 being healthy, without current or history of cardiovascular, respiratory, metabolic or

141 musculoskeletal disease, and none were smokers or taking any medications that might

142 affect the cardiovascular or hemodynamic responses to exercise. The protocol,

143 including possible risks and discomforts related to the testing and exercise training,

144 was provided to the subjects both verbally and in writing prior to commencement of

145 data collection. Participants were instructed to maintain their normal diets over the

146 course of the study. Subjects provided written informed consent before voluntary
 8

147 participation in this study. All procedures in this study were approved by The

148 University of Western Ontario Ethics Committee for Research on Human Subjects.

149

150 Experimental Protocol

151 High-Intensity Training (HIT) Protocol

152 The HIT program used in the present study was slightly modified from that

153 used previously in our laboratory (41). All training was completed on a friction-

154 braked cycle ergometer (Monark Ergomedic 874E, Monark, Vansbro, Sweden), with

155 an investigator always present. Each HIT session began with 5 min of “loadless”

156 cycling, with no external resistance applied. Then, training participants cycled for 1

157 min at 110% of the WRmax attained during the maximal ramp incremental (RI) test

158 (see below). These work intervals were followed by 1 min of “loadless” cycling.

159 Work intervals were repeated 8 times during the first training session, progressing to

160 a total of 12 intervals by the final (twelfth) training session; each training session was

161 separated by 1-2 days recovery (Fig. 1). Participants cycled at a self-selected cadence

162 between 80-100 rpm, and maintained that cadence across all work intervals of the

163 training sessions. Total work and adherence to the selected cadence were monitored

164 by the investigator and recorded for each HIT session. After every two training

165 sessions, the cycling load was increased by ~2-4%, and an additional training interval

166 was added (Fig. 1), in order to promote continuous improvements in training

167 performance. During training sessions, participants were provided with strong verbal

168 encouragement, were allowed water ad libitum, and remained on the cycle ergometer

169 for the entire duration of the session.

 9

170 Exercise Testing

171 At the start of the study (pre-training, PRE), all participants visited the

172 laboratory for an initial ramp incremental (RI) plus constant-load, step exercise (SE)

173  O 2max ) and

test (RISE-105) to volitional fatigue to determine their maximal V O 2p ( V

174 estimated lactate threshold ( θ̂ L ) (50). Over the following two visits, they completed

175 step-transitions in work rate (WR) from a baseline of 20 W to a final intensity

176 representing 90% θ̂ L (i.e., within the moderate-intensity exercise domain; MOD)

177 performed either as a i) single-step protocol (single step, 20 W→90%), or ii) as a

178 double-step protocol where the increases in WR were performed as two identical

179 step-transition increments in WR from 20 W to ~ 45% θ̂ L (lower step, LS), followed

180 by a step-transition from ~ 45% θ̂ L to 90% θ̂ L (upper step, US). Each step-transition

181 lasted 6 min and each step protocol was repeated either 3 (single-step; SS) or 5 times

182 (double-step; DS). The effectiveness of the HIT program was assessed by means of a

183 time-to-fatigue (TTF) performance test consisting of a constant-load cycling test

184 performed at the same absolute WR which corresponded to the maximal WR (WRmax)

185 achieved during the PRE RI exercise test. For the HIT group, the RISE-105 exercise

186 test, the MOD step-transitions and the TTF test were repeated following 6 (after ~2

187 weeks; MID) and 12 training sessions (after ~4 weeks; POST). The CON group

188 performed identical sets of testing separated by 2- (MID) and 4-weeks (POST). For

189 both the HIT and CON groups, a final “verification” (VER) testing session was

190 performed 4-5 days following POST testing and consisted of a RISE-105 test, single-

191 step MOD transitions (total of 3 transitions), double-step MOD transitions (total of 5

192 transitions) and a TTF test. Testing and training timelines for both groups are
 10

193 illustrated in Figure 1. Testing at the PRE, MID, POST, and VER times was intended

194 to establish a time course for physiological adaptations and performance

195 enhancements over the course of the twelve HIT sessions. The MOD exercise

196 transitions were performed at VER in order to confirm that any changes in V O 2p

197 kinetics were a result of longer-term HIT training, and not a consequence of short-

198 term effects resulting from the previous (final) exercise training sessions. Participants

199 abstained from caffeine for at least 3 h, and from alcohol for at least 12 h prior to all

200 testing.

201 Ramp Incremental Exercise (RISE-105) Test

202 Participants completed a ramp incremental (RI) exercise test (20 W/min) to

203 their limit of tolerance at each of the PRE, MID, POST and VER time points. The

204 RISE-105 test was performed on an electromagnetically-braked leg cycle ergometer

205 (H-300-R Lode; Lode B.V., Groningen, Holland), and was used to determine both

206  O 2max ) and estimated lactate threshold ( θ̂ L ). The θ̂ L was

maximal O2 uptake ( V

207  CO2p ) and

determined visually as the V O 2p at which the rise in CO2 output ( V


208 ventilation ( VE ) became disproportionate to the rise in V O 2p , along with a systematic

209  E / V O 2p ) and end-tidal PO2, and where

increase in the ventilation-to- V O 2p ratio ( V

210  CO2p ratio ( V

the ventilation-to- V  CO2p ) and end-tidal PCO2 remained stable (6).
E /V

211 A value for θ̂L was identified for each of the listed criteria and values were averaged

 O 2max
212 together to attain a final estimate for θ̂L . V was determined as the average

 O 2p
V
213 measured during the final 20 s of each of the RI and SE portions of the RISE-

214 105 protocol (50): at volitional fatigue at the end of the RI test, the WRmax was noted,
 11

215 and WR was returned to 20 W. Participants continued cycling at 20 W for 5 min, after

216 which the WR was increased as a step-function to 105% WRmax (SE-105).

217 Participants cycled at this higher WR until they could no longer maintain a cadence

218 above 50 revolutions per minute (rpm) despite strong, verbal encouragement.

219 Moderate Intensity Step-Transition Tests

220 Five repetitions of the double-step MOD transition tests (DS-MOD) were

221 completed on a cycle ergometer (H-300-R Lode; Lode B.V., Groningen, Holland) at

222 each of the PRE, MID, POST and VER testing points. Additionally, 3 repetitions of

223 the single-step MOD transition tests (SS-MOD) were completed at both PRE and

224 VER time points. The SS-MOD consisted of 6 min cycling at a baseline of 20W,

225 followed by an instantaneous step-transition to a WR corresponding to 90% θ̂ L for an

226 additional 6 min. For the DS-MODs, participants cycled for 6 min at 20W, followed

227 by two 6 min step transitions; the first lower step (LS) involved an instantaneous

228 increase to a WR midway between 20W and the WR corresponding to 90% θ̂ L

229 (ΔWRLS), and the second upper step (US) involved an instantaneous increase in WR

230 from ΔWRLS to a WR corresponding to 90% θ̂ L (ΔWRUS); the change in WR for LS

231 and US were identical (i.e., ΔWRLS = ΔWRUS). Transitions lasted 6 min to allow the

232 achievement of steady-state V O 2p . Participants maintained a cadence of ~ 70 rpm

233 during these tests. Because HIT was performed on alternate days, it was not possible

234 to perform only single transitions per day, therefore multiple transitions were

235 performed on the same day, with each MOD transition separated by 10 min of resting

236 recovery. It has been previously shown that parameter estimates (time constant, time

237 delay, amplitude) for V O 2p and deoxygenation ([HHb]) kinetics do not differ when
 12

238 transitions to MOD are completed either on separate occasions, or sequentially as a

239 series of transitions (2 or 6), each separated by 6 min baseline cycling (56).

240 Time-to-Fatigue (TTF) Performance Test

241 Participants completed a constant-load TTF performance test on a cycle

242 ergometer (H-300-R Lode; Lode B.V., Groningen, Holland) at each of the PRE, MID,

243 POST and VER testing points. The test involved an initial 5 min of cycling at 20 W,

244 followed by an instantaneous increase in WR to 100% of the WRmax attained during

245 the PRE RI test. Participants were instructed to maintain a cadence of at least 70 rpm;

246 the test was stopped when subjects could not maintain a cadence of at least 60 rpm,

247 despite strong verbal encouragement.

248 Training and Testing Timelines

249 PRE testing was completed during three visits, totaling ~ 2.5 hrs. After 1-2

250 days, the HIT group began training with 1-2 days rest between sessions. Each training

251 session required 20-30 min for completion. After completion of 6 HIT sessions (~2

252 weeks), the HIT group completed MID testing during two visits, totaling ~ 2 hrs. The

253 HIT group completed another 6 training sessions (~2 weeks), followed by POST

254 testing. Post-training tests were completed during two visits, totaling ~ 2 hrs. VER

255 testing was performed 4-5 days after the POST testing, and required ~ 2.5 hrs during

256 two visits. On average, participation in this study required 5-6 weeks for completion.

257 CON participants completed all testing along a timeline identical to the HIT

258 group, but did not complete any HIT training sessions. The CON group was asked to

259 maintain their regular recreational activity (2-3 times per week), without additional

260 training.
 13

261 Data Collection

262 Gas exchange measurements were similar to those previously described (1).

263 Briefly, inspired and expired flow rates were measured with a low dead space (90 ml)

264 bidirectional turbine (Alpha Technologies VMM 110), which was calibrated prior to

265 each test with a 3.0 L syringe. The inspired and expired gases were sampled

266 continuously (every 20 ms) at the mouth and analyzed for concentrations of O2, CO2,

267 and N2 by mass spectrometry (Innovision, AMIS 2000, Lindvedvej, Denmark)

268 following calibration with precision-analyzed gas mixtures. Changes in gas

269 concentrations were aligned with gas volumes by measuring the time delay for a

270 square-wave bolus of gas passing through the turbine to the resulting changes in

271 fractional gas concentrations (measured by the mass spectrometer). Data were

272 transferred to a computer, which aligned the concentration and volume information to

273 build a profile of each breath. Breath-by-breath alveolar gas exchange was calculated

274 using algorithms (5). Heart rate (HR) was continuously monitored by three-lead

275 electrocardiogram using PowerLab (ML132/ML880; ADInstruments, Colorado

276 Springs, CO, USA) on a separate computer.

277 Local muscle deoxygenation ([HHb]) of the quadriceps vastus lateralis muscle

278 was measured using near-infrared spectroscopy (NIRS) (Oxiplex TS, Model 95205,

279 ISS, Champaign, IL, USA). The rigid sensor was placed on the belly of the muscle,

280 midway between the lateral epicondyle and greater trochanter of the femur. Emitter

281 fibers and detector were housed in a rigid, plastic sensor casing, ensuring that their

282 positions remained fixed. The sensor was clipped on to a velcro strap, which was

283 wrapped around the participant’s leg to secure its position. Additionally, an optically-
 14

284 dense black vinyl sheet was placed overtop and around the sensor casing (minimizing

285 the intrusion of extraneous light and loss of NIR light). The thigh, with attached

286 sensor and covering, was wrapped with an elastic bandage to minimize any

287 movement of the sensor.

288 The theory of tissue spectroscopy has been previously described by Elwell

289 (20). Briefly, the Oxiplex TS—providing multidistance frequency-domain

290 spectroscopy (fd-NIRS)—provided a single channel of eight laser diode light sources

291 operating at two different wavelengths (690 nm and 828 nm, four per wavelength).

292 Light from the diodes was coupled to fibers in a photomultiplier tube, and pulsed in

293 rapid succession (110 MHz). The diodes were set at source-detector distances of 2.0,

294 2.5, 3.0 and 3.5 cm. The light was received by another detector fiber, then carried

295 back and detected by a photodiode in the spectrometer. After the rigid sensor was

296 secured on the leg, detector gain was adjusted for an optimal signal as the subject

297 rested on the cycle ergometer (detector “gain” is dependent on detector bias voltage,

298 where a larger voltage produces a larger signal). The OxiplexTS produced 25

299 measurements per second; averaged measurements were displayed and recorded at a

300 frequency of 1 Hz. The concentration of NIRS-derived muscle deoxyhaemoglobin

301 (HHb) is presented as arbitrary units (AU).

302 Data Analysis

303 Breath-by-breath V O 2p data were filtered by removing aberrant data points

304 that lay outside 4 SD of the local mean (37, 50). The remaining data were interpolated

305 to 1 s intervals, and time-aligned such that time “zero” represented the onset of the

306 MOD exercise transition (in case of the DS-MODs, the LS step transition). The data
 15

307 from test repeats were ensemble-averaged, and further time-averaged into 5 s bins to

308 yield a single profile for each subject at each testing period. The phase I-phase II

309 transition was identified as previously described (29, 52). On-transient phase II V O 2p

310 kinetics were modeled using the following equation:

311 Y(t) = YBSLN + Amp (1 – e-(t-TD)/t) (Eq. 1)


312 Where Y(t) represents V O 2p at any time (t); YBSLN is the average steady-state V O 2p

313 measured during the period immediately before the change in WR; Amp (amplitude)

314 is the steady-state increase in V O 2p above the baseline V O 2p ; τ (time constant)

315 represents the time required to attain 63% of the steady-state amplitude; and TD is the

316 time delay (mathematically generated as the point at which the exponential model is

317 predicted to intersect the baseline). Steady-state baseline V O 2p was established from

318 data 60 s prior to each change in WR. Data were modeled from the phase I-phase II

319 transition to the end of the 6 min exercise transition using Origin data fitting software

320 (OriginLab). The 95% confidence interval (CI95) for the estimated time constant was

321 determined following a preliminary fit with YBSLN, Amp and TD constrained to best-

322 fit values, with the τ allowed to vary. The mean response time (MRT) (39) of V O 2p

323 described the overall time course of V O 2p during the exercise transition and was

324 estimated using the function described in Eq. 1, but with inclusion of all V O 2p data

325 from the onset of exercise, and the TD constrained to 0 s. This approach allowed for

326 an estimate of the O2 deficit (52) for each WR transition. The O2 deficit provides

327 information on non-oxidative energy transfer, and was calculated as:

328 O2 deficit (L) = MRT (s) x Δ V O2ss (L/min) x min/60 s (Eq. 2)

 16

329 The functional gain of the fundamental V O2 response was calculated as Δ V O2ss / ΔWR

330 (mL/min/W).

331 NIRS-derived data were time-aligned and ensemble-averaged into 5 s bins to

332 yield a single response time for each subject. The time-course of adjustment for

333 [HHb] has been described to consist of a TD following the onset of exercise, with a

334 subsequent “exponential-like” increase in the signal with time of exercise (17). The

335 TD for the [HHb] ([HHb]TD) response was determined visually using second-by-

336 second data, and was identified as time between the step-increase in WR and the time

337 where the [HHb] signal began to rise systematically above the nadir in the signal

338 response (28). [HHb]TD was determined for individual trials, and averaged for each

339 LS, US and SS, at each testing point, for each subject. The ensemble-averaged [HHb]

340 responses were modeled from [HHb]TD to 90 s of the transition, with a

341 monoexponential function of the form in Eq. 1 to determine the time course of muscle

342 [HHb] (τ[HHb]). A systematic decline in [HHb] did not occur prior to 90 s of the

343 transition in any subjects. Baseline [HHb] ([HHb]BSLN) was determined for each of

344 the US and LS as the mean value in the 60 s prior to a transition. The effective time

345 constant (τ’ = [HHb]TD + τ[HHb]) was calculated to describe the overall time course

346 for muscle [HHb]. The steady-state value for [HHb] was determined as the end point

347 for the fitting of the monoexponential function ([HHb]ss). The steady-state increase in

348 muscle deoxygenation in relation to the change in muscle O2 utilization was

349 calculated as Δ[HHb]ss / Δ V O2p ss.

350 Second-by-second [HHb] and V O 2p data were normalized for each subject (0-

351 100% of response). The normalized phase II V O 2p data were modeled (from 0-100%
 17

352 of response) using the time constants for individual subjects for LS, US and SS

353 transitions. The normalized adjustment of muscle O2 utilization was estimated by


354 shifting the normalized phase II V O 2p response profile towards the start of each of

355 the LS, US and SS step-transitions (i.e., t = 0 for the LS and SS, and t = 360 s for the

356 US) by a time corresponding to the estimated phase II TD for each transition, thereby


357 making the normalized Δ V O 2p at the immediate onset of the transition equal to

358 “zero”. Data were further averaged into 5 s bins and a Δ[HHb]-to-Δ V O2p ratio was

359 calculated for each of the LS, US and SS exercise transitions, with a value of 1.0

360 corresponding to “steady-state” conditions for each of the Δ[HHb] and Δ V O2p

361 variables. An “overshoot” in the Δ[HHb]-to-Δ V O2p response profile was estimated

362 by integrating the area bounded by the Δ[HHb]-to-Δ V O2p profile and a ratio value

363 equal to 1.0. The area was determined for the period 20-180 s of each transition, at

364 each testing point for each subject, as this time window incorporated data beyond the

365 NIRS-derived [HHb]TD, and was of sufficient duration for the V O2p and [HHb]

366 signals to reach 100% of their steady-state amplitudes.

367 Statistical Analysis

368 Data are presented as means ± SD. Repeated measures analysis of variance

369 (ANOVA) was used to determine statistical significance for the dependent variables;

370 repeated factor of time (PRE, MID, POST and VER), and between factor of group

371 (HIT, CON). Where group x time interactions were identified, a Tukey post-hoc

372 analysis was used to identify differences between conditions. ANOVA was analyzed
 18

373 using SPSS Version 17.0 (SPSS Inc., Chicago, IL). Statistical significance was

374 accepted at P < 0.05.

375 Results

376 Exercise Performance (TTF and Training)

377 Relative training intensity remained constant across the 12 training sessions (~

378  O 2max ), but training interval power output and exercise volume were
110% V

379 increased (P < 0.05) during the final 6 training sessions (Table 1).

380 Exercise performance, measured as the time-to-fatigue (TTF) and peak WR

381 during the RISE-105 test, increased (P < 0.05) with HIT over the course of the 12

382 training sessions, while CON group showed no changes (Table 2). The TTF when

383 exercising at the constant WRmax achieved in the initial (PRE) RI test increased by

384 85% (P < 0.01) from PRE to VER, while the WRmax achieved during RI testing

385 increased by ~ 17% (P < 0.01) during this same period. Improvements for both TTF

386 and RI WRmax were observed from PRE to MID (P < 0.05), and MID to POST (P <

387 0.05); no differences occurred between POST and VER. There were no differences

388 present at baseline (PRE) TTF or WRmax between the HIT and CON groups.

389 Maximal O2 uptake and Lactate Threshold

390  O 2max increased by 6% (P < 0.05) from PRE to

In the HIT group, absolute V

391  O 2max increased by

MID, and by 11% (P<0.05) from MID to VER, while relative V

392 7% and 9%, respectively, for these same training periods (P < 0.05) (Table 2);

393  O 2max remained unchanged in CON at all testing times.

V
 19

394 In the HIT group, the WR and V O 2p corresponding to the estimated lactate

395 threshold ( θ̂L ) increased from PRE to MID (P < 0.05), and MID to POST (P < 0.05),

396 with no changes POST to VER (see Table 2); the increases from PRE to POST

397 training were 20% (P < 0.01) and 35% (P < 0.01), respectively, relative to V O 2p and

398 WR (Table 2). The θ̂L remained unchanged in CON throughout the study. The θ̂L was

399 not different between groups at the start of training.

400 During PRE, the end-step, steady-state V O 2p for the LS and US represented

401  O 2max . In the US, the slope

72% and 98% of PRE θ̂L , and 42% and 58% of the PRE V


402 of the V O 2p -time relationship during the latter part of the transition was not different

403 from “zero” indicating the attainment of steady-state conditions with no evidence of a

404 “slow-component”. This confirms that the exercise intensity was confined within the

405 MOD domain during the US.

406  O 2p kinetics
V

407 Table 3 outlines V O 2p kinetics parameters for both HIT and CON groups, at

408 PRE, MID, POST and VER testing points. Figure 2 shows the group mean, ensemble-

409 averaged V O 2p response profiles for the double (2A) and single (2B) step-transitions

410 within the moderate-intensity domain for the HIT group at PRE and VER. Prior to the

411 start of training there were no differences between the CON and HIT groups for any

412 of the V O 2p parameter estimates (τ V O 2p , V O 2p ss, V O 2p bsl, V O 2p amplitude, TD),

413 the V O 2p gain (Δ V O 2p ss /ΔWR) or O2 deficit. Changes in LS or US V O 2p kinetics

414 parameters were not seen in CON across testing periods.

 20

415 Upper Step vs. Lower Step. Despite the same ΔWR in the LS and US, the τ

416  O 2p (P < 0.05) and V

V  O 2p gain (P < 0.01) were greater in the US than in the LS at

417 all time points for both the HIT and CON groups. In the HIT group, the O2 deficit was

418 ~ 60% greater in the US than in the LS at PRE (423 ± 125 mL vs. 265 ± 58 mL,

419 respectively; P < 0.01). At VER, however, the O2 deficit was greater in the US than in

420 the LS (303 ± 100 mL vs. 246 ± 73 mL), consequent to a greater reduction (P < 0.05)

421 of the O2 deficit in the US. In the CON group, the O2 deficit was greater in US than

422 in LS and there were no changes between time points.

423 Lower Step. In the HIT group, the τ V O 2p was reduced by 38% from PRE to

424 POST (24 ± 6 s and 15 ± 2 s, respectively; P < 0.01). No differences in τ V O 2p

425 occurred between POST (15 ± 2 s) and VER (14 ± 3 s). TD in the HIT group

426 increased from PRE to POST (13 ± 4 s and 20 ± 2 s, respectively; P < 0.05), with no

427 differences occurring between POST and VER. No changes in V O 2p ss, V O 2p bsl,

428  O 2p amplitude, V
V  O 2p gain or O2 deficit occurred as a consequence of the training

429 program. No significant changes occurred over time in CON subjects.

430 Upper Step. In the HIT group, the τ V O 2p was reduced by 38% from PRE to

431 POST (45 ± 5 s and 28 ± 7 s, respectively; P < 0.01). No change in τ V O 2p occurred

432 between POST and VER. The HIT group TD increased from PRE to POST (5 ± 10 s

433 and 11 ± 6 s, respectively; P < 0.05), with no differences occurring between POST

434 and VER. O2 deficit was reduced by 22% from PRE to POST (423 ± 125 mL and 329

435 ± 76 mL, respectively; P < 0.05), with no differences between POST and VER. No

436 changes in V O 2p ss, V O 2p bsl, V O 2p amplitude or V O 2p gain occurred as a

 21

437 consequence of the training program. There were no changes in the CON group over

438 time.

439 Single Step. In the HIT group, the τ V O 2p was reduced by 41% from PRE to

440 VER (32 ± 7 s and 19 ± 3 s, respectively; P < 0.01). The O2 deficit decreased by 20%

441 in the HIT group (661 ± 110 mL and 527 ± 102 mL, respectively; P < 0.01). There

442 were no changes in V O 2p ss, V O 2p bsl, V O 2p amplitude, or V O 2p gain. In the CON

443 group, τ V O 2p , V O 2p gain and O2 deficit remained unchanged over time; however,

444 both V O 2p bsl and V O 2p ss were reduced (P < 0.05) at VER.

445 NIRS-derived [HHb] kinetics and muscle oxygenation parameters

446 Table 4 shows [HHb] kinetics parameters for HIT and CON groups, at PRE,

447 MID, POST and VER time points. The effective time constant (τ’[HHb]) was

448 consistently larger in the US compared to the LS, in both the HIT and CON groups (P

449 < 0.05 at all time points). Despite a significant lowering of τ V O 2p following 12

450 sessions of HIT, there was no change in the adjustment of [HHb] (τ[HHb] and

451 τ’[HHb]) in the LS, US and SS transitions. The CON group also did not experience

452 any changes over time in τ[HHb] and τ’[HHb]. In both the HIT and CON groups

453 there were no changes in [HHb] kinetics parameters ([HHb]bsl, [HHb]ss, [HHb] end-

454 step, and ΔHHbss/Δ V O 2p ss) for LS, US or SS at any of the training-related testing

455 times.

456 Figure 3 shows ensemble-averaged absolute and normalized [HHb] responses

457 to MOD step-transitions (LS, US and SS) in the HIT group, at PRE and VER. As

458 evidenced in Fig. 3A the [HHb] profile exhibited a transient overshoot during the LS

459 prior to testing that was not seen during VER. This overshoot was observed in 6 of 8
 22

460 HIT subjects and was not evident at MID, POST or VER; in CON the overshoot was

461 observed in 3 of 5 subjects at both PRE and VER.

462 Normalized Δ[HHb] / ΔVO2 ratio

463 Figure 4 shows the mean profiles for Δ[HHb] / Δ V O 2p ratios in LS and US

464 for the HIT group. At PRE, the Δ[HHb] / Δ V O 2p ratios for LS and US were 1.05 and

465 1.20, respectively (both significantly different from 1.0, P < 0.05); by MID this

466 “overshoot” was no longer apparent in either the LS or US (i.e. the ratio was no

467 longer different from 1.0). These ratios remained at ~1.0 throughout the remainder of

468 the study (MID to VER), consistent with steady-state conditions for both variables.

469 Similarly, Figure 5 illustrates mean profile for Δ[HHb] / Δ V O 2p ratio during SS at

470 PRE and VER in the HIT group. Prior to training (PRE), the overall ratio from 20 to

471 180 s of the transition averaged 1.09 (significantly different from 1.0, P < 0.05);

472 following training (VER), no overshoot was observed.

473 Discussion

474 This study investigated the effects of 12 sessions of high-intensity interval

475 training (HIT) on phase II pulmonary O2 uptake ( V O 2p ) and muscle deoxygenation

476 ([HHb]) kinetics during transitions to constant-load, moderate intensity (MOD)

477 exercise initiated from a lower (LS) and higher (US) baseline WR and metabolic rate.

478 In agreement with previous reports (9, 32, 40, 55), we observed that for similar

479 increments in work rate (ΔWR), there was a slower adjustment of both V O 2p and

480 [HHb] in the US compared to the LS, as well as a greater steady-state O2 cost of

481 exercise (gain, ∆ V O 2p /∆WR) in the US, reflecting lower metabolic efficiency. As
 23

482 hypothesized, there was a speeding of V O 2p kinetics in both the LS and US following

483 4 weeks (i.e., 12 sessions) of HIT training. The absolute reduction of τ V O 2p in the

484 US (~20 s) was greater than that seen in the LS (~10 s); however, these changes

485 represented a similar relative speeding in both steps (~40%). Despite the training-

486 induced speeding of V O 2p , the adjustment of muscle deoxygenation ([HHb]),

487 reflecting fractional O2 extraction and the ratio of microvascular blood flow-to-muscle

488 O2 utilization, did not change. The observed speeding of V O 2p kinetics without

489 changes in the adjustment muscle deoxygenation agree with the findings of McKay et

490 al. (41), who reported a faster adjustment of V O 2p in transitions from 20 W to a WR

491 corresponding to 90% θ̂ L (performed as a single step). Overall, these findings suggest

492 that the training-induced adjustments of V O 2p (and muscle O2 utilization) are

493 independent of the baseline metabolic rate, and also that microvascular blood flow

494 kinetics matched those of muscle O2 utilization. To our knowledge, this is the first

495 study to observe training-induced speeding of V O 2p kinetics initiated from a lower

496 and higher baseline metabolic rate within the MOD domain.

497 Effect of HIT on V O 2p and [HHb] kinetics

498 Following 12 sessions of HIT, τ V O 2p was reduced by ~40% when the

499 exercise transition was performed as a single step (SS) of 20 W to ~90% θ̂L ; τ V O 2p

500 decreased from 32 s to 19 s. These changes in τ V O 2p mirror those seen by McKay et

501 al. (41), where 8 sessions of HIT resulted in a 12 s reduction in τ V O 2p during SS

502 transitions. In agreement with McKay et al. (41), the V O 2p gain was not affected by

503 training, remaining at ~9-10 mL/min/W. As a consequence of similar V O 2p gain but

 24

504 faster V O 2p kinetics, the O2 deficit was reduced from 661 ml (PRE) to 527 ml (VER).

505 In agreement, Bailey et al. (3) and Berger et al. (7) used a short-term interval training

506 model and reported reductions in SS-MOD τ V O 2p from 28 s to 21 s (3) and 32 s to

507 21s (7).

508 As previously reported (9, 32, 40, 55), the adjustment of V O 2p was slower

509 and the V O 2p gain was greater when the exercise transition was initiated from an

510 elevated baseline metabolic rate (US vs LS). There was a speeding of V O 2p kinetics

511 following 12 sessions of HIT during transitions in both LS and US, as τ V O 2p

512 decreased by ~ 20 s (US) and ~ 10 s (LS). Several factors have been suggested to

513 contribute to the slowed adjustment of V O 2p within the upper region of the MOD

514 domain: O2 availability to the working muscle (32, 40), hierarchical recruitment of

515 additional muscle fibres (9), and intramuscular energetic state (8). Prior work has

516 described slowed adjustments of femoral artery (bulk) blood flow and O2 delivery in

517 the US compared to LS, suggesting that the greater τ V O 2p in the US occurred as a

518 consequence of a slowed adjustment of bulk O2 delivery (32). Recently Bowen et al.

519 (8) suggested that when exercise transitions are initiated from an elevated baseline

520 metabolic rate, factors other than slowed blood flow adjustment and intrinsic

521 metabolic properties of newly recruited muscle fibres could account for the slower

522 adjustment of V O 2p in the US compared to LS. They proposed that during

523 transitions from a raised metabolic rate, a lowered metabolic energy state (i.e.,

524 elevated muscle concentrations of free ADP and Pi; lower ATP) and lowered free

525 energy release from ATP hydrolysis (“less negative” ∆GATP) would require greater

526 activation of mitochondrial oxidative energy production and muscle O2 utilization,

 25

527 compared to when exercise was initiated from a lower metabolic rate (8, 24). While

528 this hypothesis remains to be tested, the findings do suggest that slowed V O 2p

529 kinetics in the upper region of the MOD domain likely are a consequence of

530 limitations imposed by a combination of adjustments to metabolism, blood flow and

531 O2 delivery.

532 Despite the faster HIT-induced adjustments in V O 2p (and muscle O2

533 utilization) in the LS, US and SS, the overall adjustment (TD + τ) of muscle

534 deoxygenation ([HHb]) was unchanged over the course of the 12 HIT sessions. These

535 results agree with previous studies from our laboratory (41, 44), but differ from those

536 of Bailey and coworkers (3) who reported faster kinetics for both V O 2p and muscle

537 deoxygenation during transitions into MOD after repeated sprint training. In the

538 present study, the speeding of the V O 2p kinetics without accompanying changes in

539 muscle deoxygenation kinetics suggests that the adjustment of microvascular blood

540 flow also became faster with training. Consistent with this is the transient overshoot

541 in the profile of the normalized ∆[HHb]/∆ V O 2p ratio, which reflects a transient

542 decrease in microvascular blood flow-to-muscle O2 utilization and increased

543 fractional muscle O2 extraction. This overshoot was present in the exercise transition

544 at PRE, but not seen at MID, POST and VER. Using a lower intensity END training

545 program, Murias et al. (43) reported a reduction in the normalized ∆[HHb]/∆ V O 2p

546 overshoot after 3 wks training, and its eliminatin after 6 wks training, with these

547 changes being closely associated with reductions in τ V O 2p (see Fig. 4 in (43)). In the

548 present study, however, V O 2p kinetics continued to speed over the entire twelve HIT

549 sessions while the reduced ∆[HHb]/∆ V O 2p overshoot was seen only between PRE
 26

550 and MID. These observations suggest that training-induced adaptations contributing

551 to the elimination of an overshoot do not entirely account for the speeding of V O 2p

552 kinetics seen throughout the HIT program. Possible mechanisms for the speeding of

553 V O 2p kinetics are discussed below.

554 Potential mechanisms contributing to speeding of V O 2p kinetics

555 The current study is, to our knowledge, the first to examine training

556 adaptations in separate lower and upper regions of the MOD domain, where

557 significant differences in V O 2p kinetics and gain (present study; 8, 9, 40), heart rate

558 kinetics (8, 40), conduit artery blood flow and vascular conductance kinetics (40),

559 muscle deoxygenation kinetics (present study; 8, 40), and O2 deficit (present study; 9)

560 have been observed. While the speeding of V O 2p kinetics in both LS and US, without

561 changes in muscle deoxygenation kinetics, suggests an improved dynamic matching

562 of microvascular O2 delivery with muscle O2 utilization, markers of microvascular

563 and muscle metabolic function were not directly measured in this study. Thus, we

564 can only speculate on the specific physiological adaptations at the level of the muscle

565 that may have contributed to the speeding of V O 2p kinetics.

566 Training adaptations in the adjustment of microvascular blood flow likely

567 contributed, in part, to the speeding of V O 2p kinetics, especially early in the training

568 period (between the PRE and MID), as evidenced by the elimination of a ∆[HHb]/∆

569  O 2p overshoot by MID. However, adaptations to microvascular blood flow

V

570 dynamics likely continued with training because as τ V O 2p continued to fall

571 significantly in both LS (by 22%) and US (by 29%) from MID to VER (Table 3A),
 27

572 the τ’[HHb] did not change significantly (Table 4A). This suggests that the dynamic

573 relationship between microvacular blood flow and muscle O2 utilization, and thus

574 muscle deoxygenation, was maintained. Green et al. (26) reported increases in muscle

575 capillarity with a 6-day training model (at ~ 67% VO2peak for 2 hr/day): increases in

576 capillary contacts per muscle fibre area and capillary density (without changes in

577 muscle oxidative potential) were reported after 3 training days, while increases in

578 capillary contacts per muscle fibre, capillary contacts per muscle fibre area and

579 capillary density were reported after 6 training days. These increases in muscle

580 capillarity would be expected to increase the functional capillary surface area, and

581 thus the potential for microvascular convective and diffusive O2 delivery to the

582 muscle mitochondria. Using a similar training program Shoemaker et al. (54) reported

583 faster leg conduit artery blood velocity and vascular conductance kinetics after 10

584 training days, while Phillips et al. (44) reported faster V O 2p kinetics after 4 training

585 days, prior to any increases in muscle oxidative potential.

586 In the present study, the continued speeding of V O 2p kinetics seen late in the

587 HIT program likely suggest that mechanisms in addition to microvascular blood flow

588 adaptations may have gradually contributed to the training adaptations (between MID

589 and VER). Using an intense, intermittent sprint-training protocol, Gibala and

590 coworkers (10, 12, 22, 23) observed increases in maximal activities of cytochrome

591 oxidase (COX) (22), citrate synthase (12) and β-hydroxyacyl CoA dehydrogenase (β-

592 HAD) (12); pyruvate dehydrogenase (PDH) (12), COX II and COX IV (10, 22)

593 protein content; peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α)

594 mRNA (23) and protein content (12); and phosphorylation of AMP kinase (AMPK)
 28

595 (22), p38 mitogen-activated protein kinase (p38-MAPK) (22) and

596 calcium/calmodulin-dependent protein kinase (CaMK) (22). Altogether, these

597 alterations in these markers reflect increases in cell signaling leading to mitochondrial

598 biogenesis and increases in muscle oxidative capacity, after only 1-6 training

599 sessions. Given the HIT program used in the present study, similar training-induced

600 adaptations might be expected, although the somewhat lower training intensity used

601 in the present study (repeated exercise bouts at 110% of pre-training maximal work

602 rate versus repeated “all-out” exercise bouts) may have induced a differing rate of

603 metabolic adaptation. A higher oxidative capacity of the trained muscle, thus, may

604 have contributed to the continued speeding of V O 2p kinetics seen in the present

605 study.

606  O 2max , estimated lactate threshold and exercise performance

Effect of HIT on V

607  O 2max and exercise performance (TTF and WRmax)

Improvements in V

608 occurred following 12 sessions of HIT, demonstrating the effectiveness of this

609 training program in improving training status and overall fitness. In the present study,

610  O 2max (relative and absolute), WRmax in the RI test, and the WR
improvements in V

611 and VO2p corresponding to θ̂L , occurred from PRE to MID testing, and further from

612 MID to POST. The greatest improvements occurred in TTF, with an 85% increase

613 (209 s to 386 s) following 12 sessions of training. These findings agree with the

614 findings of McKay et al. (41), whose similar HIT protocol promoted improvements in

615  O 2max , WRmax during RI testing and the WR and VO2p

relative (but not absolute) V

616 corresponding to θ̂L . Additionally, several investigations employing short-term HIT

 29

617 (2-week (3, 22) and 6-week (10) programs) have reported improvements in exercise

618 performance, with reduced time to complete time-trials (10, 22), increased power

619 output during the time trail (22), and greater time to fatigue during a severe-intensity

620 exercise test (3).

621  O 2max (both relative and absolute), exercise tolerance

The improvements in V

622 and performance in the current study likely occurred, at least in part, due to improved

623 muscle oxidative capacity, faster adjustment of mitochondrial oxidative metabolism

624 and reduced reliance on substrate-level phosphorylation (possibly with slower rates of

625 depletion of “anaerobic” energy stores, glycogen and PCr, and slower rates of

626 accumulation of free ADP, Pi and H+, and maintenance of a higher ∆GATP), enhanced

627 buffering and transport of fatigue-inducing metabolites. Increases in resting muscle

628 glycogen and glycolytic capacity have also been associated with HIT, and these

629 adaptations may also have contributed to the improvements in performance (11).

630 Limitations

631 In the present study, there is some uncertainty when using the noninvasive measure of

632 phase II V O 2p profile as a proxy for muscle O2 utilization in the calculation of the

633 Δ[HHb] /Δ V O 2p ratio. It is possible that by eliminating the phase I V O 2p response

634 and focusing only on the phase II response that this approach may not reflect the

635 actual response at the muscle. Krustrup and colleagues (35) reported that after the

636 onset of exercise there is a shortening of the muscle capillary-to-lung transit time

637 resulting in deoxygenated blood from the exercising muscle appearing at the lung

638 after a delay of ~ 10-15 s. They concluded, however, that despite the early arrival of

639 deoxygenated blood from contracting muscle, its contribution to the V O 2p response
 30

640 was “quantitatively small” and did not distort the fidelity of the relationship between

641 muscle O2 uptake and V O 2p kinetics. Additionally, it is reasonable to assume that the

642 increase in muscle ATP turnover and initial activation and increase in the rate of

643 muscle O2 utilization (and V O 2p ) should coincide with, or occur very soon after, the

644 transition to a higher WR has taken place (i.e., with minimal TD) (for review see

645 (51)), and that the increase in muscle O2 utilization would be “exponential-like” (with

646 a time course reflecting the phase II V O 2p response) (25, 35, 51). Therefore, we

647 believe that this adjustment then provides a reasonable estimation of events occurring

648 in the recruited muscle at the immediate onset of the exercise transition.

649 Additionally, direct measures of HIT-induced adaptations in the muscle were

650 not made and we can only speculate on adaptations based on published literature.

651 However, given the low volume, high-intensity nature of the HIT program used in

652 this study was modeled after Gibala and coworkers ((10-13, 22, 23) where muscle

653 biopsy data are presented, we are confident that similar muscle adaptations will occur.

654 These adaptations require further study.

655 Conclusions

656 In summary, the results of the present study demonstrated that low volume,

657 high-intensity interval training (HIT: 12 training sessions over 4 weeks; 8-12 intervals

658 per session at 110% VO2max; 60 s exercise: 60 s recovery) was associated with a

659 speeding of V O 2p kinetics during transitions into the MOD exercise domain,

660 regardless of whether the exercise transition was initiated from a low or elevated

661 baseline metabolic rate. Also, muscle deoxygenation kinetics were unchanged

662 despite the faster rate adjustment of V O 2p . While the relative speeding of the LS and
 31

663 US did not differ, the absolute speeding of the US was twice that of the LS following

664 the training program. Speeding of V O 2p kinetics in both the US and LS occurred

665 without parallel changes in the adjustment of muscle deoxygenation, suggesting a role

666 for improved coordination of microvascular blood flow with O2 utilization,

667 particularly in the early stages of training, with training-induced adaptations in

668 muscle oxidative capacity likely contributing more with the duration of HIT. While

669 the specific physiological mechanisms controlling the adjustment of O2 uptake for

670 each of the lower and upper regions of the MOD exercise domain remain elusive, a

671 prominent body of evidence suggests that training-induced improvements in τ V O 2p

672 for LS and US transitions may involve metabolic remodeling of the active muscle.

673 Acknowledgments

674 We thank Brad Hansen for technical support, as well as all the subjects who

675 participated in this study. We would also like to acknowledge the assistance provided

676 by Juan Murias and Matt Spencer. This study was supported by grants from the

677 Natural Sciences and Engineering Research Council of Canada. Additional support

678 was provided by grants from The University of Western Ontario Academic

679 Development Fund, the Canadian Foundation for Innovation, and Ontario Innovation

680 Trust. Financial support to A. M. Williams was provided by an Ontario Graduate

681 Scholarship.

682
 32

683 References
684
685 1. Babcock MA, Paterson DH, and Cunningham DA. Effects of aerobic

686 endurance training on gas exchange kinetics of older men. Med Sci Sports Exerc 26:

687 447-452, 1994.

688 2. Bailey SJ, Vanhatalo A, Wilkerson DP, Dimenna FJ, and Jones AM.

689 Optimizing the "priming" effect: influence of prior exercise intensity and recovery

690 duration on O2 uptake kinetics and severe-intensity exercise tolerance. J Appl Physiol

691 107: 1743-1756, 2009.

692 3. Bailey SJ, Wilkerson DP, DiMenna FJ, and Jones AM. Influence of

693 repeated sprint training on pulmonary O2 uptake and muscle deoxygenation kinetics

694 in humans. J Appl Physiol 106: 1875-1887, 2009.

695 4. Barstow TJ, Lamarra N, and Whipp BJ. Modulation of muscle and

696 pulmonary O2 uptakes by circulatory dynamics during exercise. J Appl Physiol 68:

697 979-989, 1990.

698 5. Beaver WL, Lamarra N, and Wasserman K. Breath-by-breath

699 measurement of true alveolar gas exchange. J Appl Physiol 51: 1662-1675, 1981.

700 6. Beaver WL, Wasserman K, and Whipp BJ. A new method for detecting

701 anaerobic threshold by gas exchange. J Appl Physiol 60: 2020-2027, 1986.

702 7. Berger NJ, Tolfrey K, Williams AG, and Jones AM. Influence of

703 continuous and interval training on oxygen uptake on-kinetics. Med Sci Sports Exerc

704 38: 504-512, 2006.

705 8. Bowen TS, Murgatroyd SR, Cannon DT, Cuff TJ, Lainey AF,

706 Marjerrison AD, Spencer MD, Benson AP, Paterson DH, Kowalchuk JM, and
 33

707 Rossiter HB. A raised metabolic rate slows pulmonary O2 uptake kinetics on

708 transition to moderate-intensity exercise in humans independently of work rate. Exp

709 Physiol 96: 1049-1061, 2011.

710 9. Brittain CJ, Rossiter HB, Kowalchuk JM, and Whipp BJ. Effect of prior

711 metabolic rate on the kinetics of oxygen uptake during moderate-intensity exercise.

712 Eur J Appl Physiol 86: 125-134, 2001.

713 10. Burgomaster KA, Cermak NM, Phillips SM, Benton CR, Bonen A, and

714 Gibala MJ. Divergent response of metabolite transport proteins in human skeletal

715 muscle after sprint interval training and detraining. Am J Physiol Regulatory

716 Integrative Comp Physiol 292: R1970-1976, 2007.

717 11. Burgomaster KA, Heigenhauser GJF, and Gibala MJ. Effect of short-term

718 sprint interval training on human skeletal muscle carbohydrate metabolism during

719 exercise and time-trial performance. J Appl Physiol 100: 2041-2047, 2006.

720 12. Burgomaster KA, Howarth KR, Phillips SM, Rakobowchuk M,

721 Macdonald MJ, McGee SL, and Gibala MJ. Similar metabolic adaptations during

722 exercise after low volume sprint interval and traditional endurance training in

723 humans. J Physiol 586: 151-160, 2008.

724 13. Burgomaster KA, Hughes SC, Heigenhauser GJF, Bradwell SN, and

725 Gibala MJ. Six sessions of sprint interval training increases muscle oxidative

726 potential and cycle endurance capacity in humans. J Appl Physiol 98: 1985-1990,

727 2005.
 34

728 14. Burnley M, Doust JH, Ball D, and Jones AM. Effects of prior heavy

729 exercise on VO2 kinetics during heavy exercise are related to changes in muscle

730 activity. J Appl Physiol 93: 167-174, 2002.

731 15. Coffey VG, and Hawley JA. The molecular bases of training adaptation.

732 Sports Med 37: 737-763, 2007.

733 16. Croft L, Bartlett JD, MacLaren DP, Reilly T, Evans L, Mattey DL, Nixon

734 NB, Drust B, and Morton JP. High-intensity interval training attenuates the

735 exercise-induced increase in plasma IL-6 in response to acute exercise. Appl Physiol

736 Nutr Metab 34: 1098-1107, 2009.

737 17. DeLorey DS, Kowalchuk JM, and Paterson DH. Relationship between

738 pulmonary O2 uptake kinetics and muscle deoxygenation during moderate-intensity

739 exercise. J Appl Physiol 95: 113-120, 2003.

740 18. Demarle AP, Slawinski JJ, Laffite LP, Bocquet VG, Koralsztein JP, and

741 Billat VL. Decrease of O2 deficit is a potential factor in increased time to exhaustion

742 after specific endurance training. J Appl Physiol 90: 947-953, 2001.

743 19. DiMenna FJ, Wilkerson DP, Burnley M, and Jones AM. Influence of

744 priming exercise on pulmonary O2 uptake kinetics during transitions to high-intensity

745 exercise from an elevated baseline. J Appl Physiol 105: 538-546, 2008.

746 20. Elwell CE. A Practical Users Guide to Near Infrared Spectroscopy. London:

747 Hamamatsu Photonics, 1995.

748 21. Fitts RH, and Widrick JJ. Muscle mechanics: adaptations with exercise-

749 training. Exerc Sport Sci Rev 24: 427-473, 1996.

 35

750 22. Gibala MJ, Little JP, van Essen M, Wilkin GP, Burgomaster KA, Safdar

751 A, Raha S, and Tarnopolsky MA. Short-term sprint interval versus traditional

752 endurance training: similar initial adaptations in human skeletal muscle and exercise

753 performance. J Physiol 575: 901-911, 2006.

754 23. Gibala MJ, McGee SL, Garnham AP, Howlett KF, Snow RJ, and

755 Hargreaves M. Brief intense interval exercise activates AMPK and p38 MAPK

756 signaling and increases the expression of PGC-1alpha in human skeletal muscle. J

757 Appl Physiol 106: 929-934, 2009.

758 24. Glancy B, Barstow T, and Willis WT. Linear relation between time constant

759 of oxygen uptake kinetics, total creatine, and mitochondrial content in vitro. Am J

760 Physiol Cell Physiol 294: C79-87, 2008.

761 25. Grassi B, Poole DC, Richardson RS, Knight DR, Erickson BK, and

762 Wagner PD. Muscle O2 uptake kinetics in humans: implications for metabolic

763 control. J Appl Physiol 80: 988-998, 1996.

764 26. Green HJ, Burnett M, Kollias H, Ouyang J, Smith I, and Tupling S. Can

765 increases in capillarization explain the early adaptations in metabolic regulation in

766 human muscle to short-term training? Can J Physiol Pharmacol 90: 557-566, 2012.

767 27. Gurd BJ, Peters SJ, Heigenhauser GJ, LeBlanc PJ, Doherty TJ, Paterson

768 DH, and Kowalchuk JM. Prior heavy exercise elevates pyruvate dehydrogenase

769 activity and muscle oxygenation and speeds O2 uptake kinetics during moderate

770 exercise in older adults. Am J Physiol Regul Integr Comp Physiol 297: R877-884,

771 2009.
 36

772 28. Gurd BJ, Peters SJ, Heigenhauser GJ, LeBlanc PJ, Doherty TJ, Paterson

773 DH, and Kowalchuk JM. Prior heavy exercise elevates pyruvate dehydrogenase

774 activity and speeds O2 uptake kinetics during subsequent moderate-intensity exercise

775 in healthy young adults. J Physiol 577: 985-996, 2006.

776 29. Gurd BJ, Scheuermann BW, Paterson DH, and Kowalchuk JM. Prior

777 heavy-intensity exercise speeds VO2 kinetics during moderate-intensity exercise in

778 young adults. J Appl Physiol 98: 1371-1378, 2005.

779 30. Helgerud J, Hoydal K, Wang E, Karlsen T, Berg P, Bjerkaas M,

780 Simonsen T, Helgesen C, Hjorth N, Bach R, and Hoff J. Aerobic high-intensity

781 intervals improve VO2max more than moderate training. Med Sci Sports Exerc 39:

782 665-671, 2007.

783 31. Hirai DM, Copp SW, Ferguson SK, Holdsworth CT, McCullough DJ,

784 Behnke BJ, Musch TI, and Poole DC. Exercise training and muscle microvascular

785 oxygenation: functional role of nitric oxide. J Appl Physiol 113: 557-565, 2012.

786 32. Hughson RL, and Morrissey M. Delayed kinetics of respiratory gas

787 exchange in the transition from prior exercise. J Appl Physiol Respirat Environ

788 Exercise Physiol 52: 921-929, 1982.

789 33. Krogh A, and Lindhard J. The changes in respiration at the transition from

790 work to rest. J Physiol 53: 431-439, 1920.

791 34. Krustrup P, Hellsten Y, and Bangsbo J. Intense interval training enhances

792 human skeletal muscle oxygen uptake in the initial phase of dynamic exercise at high

793 but not at low intensities. J Physiol 559: 335-345, 2004.

 37

794 35. Krustrup P, Jones AM, Wilkerson DP, Calbet JA, and Bangsbo J.

795 Muscular and pulmonary O2 uptake kinetics during moderate- and high-intensity sub-

796 maximal knee-extensor exercise in humans. J Physiol 587: 1843-1856, 2009.

797 36. Kubukeli ZN, Noakes TD, and Dennis SC. Training techniques to improve

798 endurance exercise performances. Sports Med 32: 489-509, 2002.

799 37. Lamarra N, Whipp BJ, Ward SA, and Wasserman K. Effect of interbreath

800 fluctuations on characterizing exercise gas exchange kinetics. J Appl Physiol 62:

801 2003-2012, 1987.

802 38. Lin J, Wu H, Tarr PT, Zhang CY, Wu Z, Boss O, Michael LF, Puigserver

803 P, Isotani E, Olson EN, Lowell BB, Bassel-Duby R, and Spiegelman BM.

804 Transcriptional co-activator PGC-1 alpha drives the formation of slow-twitch muscle

805 fibres. Nature 418: 797-801, 2002.

806 39. Linnarsson D. Dynamics of pulmonary gas exchange and heart rate changes

807 at start and end of exercise. Acta Physiol Scand Suppl 415: 1-68, 1974.

808 40. MacPhee SL, Shoemaker JK, Paterson DH, and Kowalchuk JM. Kinetics

809 of O2 uptake, leg blood flow, and muscle deoxygenation are slowed in the upper

810 compared with lower region of the moderate-intensity exercise domain. J Appl

811 Physiol 99: 1822-1834, 2005.

812 41. McKay BR, Paterson DH, and Kowalchuk JM. Effect of short-term high-

813 intensity interval training vs. continuous training on O2 uptake kinetics, muscle

814 deoxygenation, and exercise performance. J Appl Physiol 107: 128-138, 2009.

815 42. Meyer RA, and Foley JM. Cellular processes integrating the metabolic

816 response to exercise. In: Handbook of Physiology (Section 12): Exercise: Regulation
 38

817 and Integration of Multiple Systems, edited by Rowell LB, and Shepherd JT. New

818 York: Oxford University Press, 1996, p. 841-869.

819 43. Murias JM, Kowalchuk JM, and Paterson DH. Speeding of VO2 kinetics

820 with endurance training in old and young men is associated with improved matching

821 of local O2 delivery to muscle O2 utilization. J Appl Physiol 108: 913-922, 2010.

822 44. Murias JM, Kowalchuk JM, and Paterson DH. Time course and

823 mechanisms of adaptations in cardiorespiratory fitness with endurance training in

824 older and young men. J Appl Physiol 108: 621-627, 2010.

825 45. Murias JM, Spencer M, Kowalchuk JM, and Paterson DH. Muscle

826 deoxygenation to VO2 relationship differs in young subjects with varying tVO2. Eur

827 J Appl Physiol 111: 3107-3118, 2011.

828 46. Paterson DH, and Whipp BJ. Asymmetries of oxygen uptake transients at

829 the on- and offset of heavy exercise in humans. J Physiol 443: 575-586, 1991.

830 47. Phillips SM, Green HJ, MacDonald MJ, and Hughson RL. Progressive

831 effect of endurance training on VO2 kinetics at the onset of submaximal exercise.

832 Journal of applied physiology (Bethesda, Md : 1985) 79: 1914-1920, 1995.

833 48. Pilegaard H, Saltin B, and Neufer PD. Exercise induces transient

834 transcriptional activation of the PGC-1alpha gene in human skeletal muscle. J Physiol

835 546: 851-858, 2003.

836 49. Rakobowchuk M, Tanguay S, Burgomaster KA, Howarth KR, Gibala

837 MJ, and Macdonald MJ. Sprint interval and traditional endurance training induce

838 similar improvements in peripheral arterial stiffness and flow-mediated dilation in

839 healthy humans. Am J Physiol Regul Integr Comp Physiol 295: R236-242, 2008.
 39

840 50. Rossiter H, Kowalchuk J, and Whipp B. A test to establish maximum O2

841 uptake despite no plateau in the O2 uptake response to ramp incremental exercise. J

842 Appl Physiol 100: 764, 2006.

843 51. Rossiter HB. Exercise: Kinetic Considerations for Gas Exchange.

844 Comprehensive Physiology 1: 203-244, 2010.

845 52. Rossiter HB, Ward SA, Doyle VL, Howe FA, Griffiths JR, and Whipp BJ.

846 Inferences from pulmonary O2 uptake with respect to intramuscular

847 [phosphocreatine] kinetics during moderate exercise in humans. J Physiol 518 ( Pt 3):

848 921-932, 1999.

849 53. Scheuermann BW, Bell C, Paterson DH, Barstow TJ, and Kowalchuk

850 JM. Oxygen uptake kinetics for moderate exercise are speeded in older humans by

851 prior heavy exercise. J Appl Physiol 92: 609-616, 2002.

852 54. Shoemaker JK, Phillips SM, Green HJ, and Hughson RL. Faster femoral

853 artery blood velocity kinetics at the onset of exercise following short-term training.

854 Cardiovascular research 31: 278-286, 1996.

855 55. Spencer M, Murias JM, Kowalchuk JM, and Paterson DH. Pulmonary O2

856 uptake and muscle deoxygenation kinetics are slowed in the upper compared with

857 lower region of the moderate-intensity exercise domain in older men. Eur J Appl

858 Physiol 111: 2139-2148, 2011.

859 56. Spencer M, Murias JM, Lamb H, Kowalchuk JM, and Paterson DH. Are

860 the parameters of VO2, heart rate and muscle deoxygenation kinetics affected by

861 serial moderate-intensity exercise transitions in a single day? Eur J Appl Physiol 111:

862 591-600, 2011.

 40

863 57. Whipp BJ, and Rossiter HB. The kinetics of oxygen uptake: Physiological

864 inferences from the parameters. In: Oxygen Uptake Kinetics in Sport, Exercise and

865 Medicine, edited by Jones AM, and Poole DC. Routledge, London, UK: 2005, p. 62-

866 94.

867 58. Whipp BJ, Ward SA, Lamarra N, Davis JA, and Wasserman K.

868 Parameters of ventilatory and gas exchange dynamics during exercise. J Appl Physiol

869 52: 1506-1513, 1982.

870 59. Whipp BJ, and Wasserman K. Oxygen uptake kinetics for various

871 intensities of constant-load work. J Appl Physiol 33: 351-356, 1972.

872 60. Wilkerson DP, and Jones AM. Effects of baseline metabolic rate on

873 pulmonary O2 uptake on-kinetics during heavy-intensity exercise in humans. Respir

874 Physiol Neurobiol 156: 203-211, 2007.

875 61. Wilkerson DP, and Jones AM. Influence of initial metabolic rate on

876 pulmonary O2 uptake on-kinetics during severe intensity exercise. Respir Physiol

877 Neurobiol 152: 204-219, 2006.

878 62. Zoladz JA, Korzeniewski B, and Grassi B. Training-induced acceleration of

879 oxygen uptake kinetics in skeletal muscle: the underlying mechanisms. J Physiol

880 Pharmacol 57: 67-84, 2006.

881

882

883
 41

884 Figure Captions

885 Fig 1. Testing and training timelines for A) high-intensity interval training (HIT)
886 group and B) control (CON) group. ** completion of one RISE-105 test, one TTF
887 test, 3 single-step moderate intensity exercise tests (SS-MODs) and 5 double-step
888 moderate intensity exercise tests (DS-MODs). *completion of one RISE-105 test, one
889 TTF test and 5 DS-MODs. Boxes indicate training session; numbers indicate number
890 of high-intensity interval repetitions completed during training session
891
892 Fig 2. Ensemble-averaged breath-by-breath V O 2p responses of HIT group to
893 moderate-intensity step tests; A) relative V O 2p response to LS and US transitions in
894 double-step, constant load tests (data normalized to the end-exercise V O 2p ); and B)
895 relative V O 2p response to SS transitions (data normalized to the end-exercise V O 2p );
896 open circles () represent pre-training tests, closed circles () represent post-training
897 tests at verification.
898
899 Fig 3. Ensemble-averaged [HHb] responses of HIT group to moderate-intensity step
900 tests; A) relative [HHb] response to LS and US transitions in double-step, constant
901 load tests (data normalized to the end-exercise [HHb]); and B) relative [HHb]
902 response to SS transitions (data normalized to the end-exercise [HHb]); open circles
903 () represent pre-training tests, closed circles () represent post-training tests at
904 verification.
905
906 Fig 4. Left panel: HIT group mean profiles for adjustments of [HHb] and V O 2p
907 (modeled and averaged from individual subject time constants) in the initial 180 s of
908 LS and US moderate-intensity exercise transitions, shown for pre-training (PRE) and
909 post-training (at verification testing; VER). Adjustments shown for 5A) LS in double-
910 step, constant load tests, and 5B) US in double-step, constant load tests. Closed
911 circles () represent time points at which the relative increase in [HHb] is greater
912 than the relative increase of V O 2p . Right panel: HIT group mean profiles for the
913 relative adjustment of [HHb] / V O 2p in the initial 180 s of LS and US moderate-
914 intensity exercise transitions, shown for pre-training (PRE) and post-training (at
915 verification testing; VER). Adjustments shown for 5A) LS in double-step, constant
916 load tests, and 5B) US in double-step, constant load tests. *Significant difference of
917 overall ∆[HHb] /∆ V O 2p (20 to 180 s) from 1.0 (P < 0.05).
918
919 Fig 5. Left panel: HIT group mean profiles for adjustments of [HHb] and V O2p
920 (modeled and averaged from individual subject time constants) in the initial 180 s of
921 SS moderate-intensity exercise transitions, shown for pre-training and post-training
922 (at verification testing). Closed circles () represent time points at which the relative
923 increase in [HHb] is greater than the relative increase of V O 2p . Right panel: HIT
924 group mean profiles for the relative adjustment of [HHb] / V O 2p in the initial 180 s of
925 SS moderate-intensity exercise transitions, shown for pre-training and post-training
926 (at verification testing). *Significant difference of overall ∆iHHb] /∆ V O 2p (20 to 180
927 s) from 1.0 (P < 0.05).
928
 42

929 Table 1. Training protocol, average work and time per training session

HIT sessions 1 to 6 HIT sessions 7 to 12

Relative Intensity (%)  O 2max

110% V  O 2max
~ 110% V

Absolute Intensity (W) 315 ± 45 335 ± 44 *

168 ± 25 222 ± 30 *
Total exercise volume (kJ)
9 11
Total exercise time (min)
23 27
Total time (rest + warm-up) (min)

930  O 2max , maximal O2 uptake. Total exercise

HIT, high-intensity interval training; V
931 volume and relative intensity per session results are based on average workloads
932 sustained during work intervals, and do not include loadless cycling. *Significantly
933 different (P < 0.01) from sessions 1 to 6.
 43

934 Table 2. Training responses for aerobic and performance parameters assessed during
935 ramp incremental and time-to-fatigue testing

936

Parameter PRE MID POST VER

HIT
 O 2max , L/min
V 3.41 ± 0.54 3.61 ± 0.64 * 3.88 ± 0.69*† 4.00 ± 0.66*†
 O 2max , mLkg min 43 ± 4
V -1 -1
46 ± 6 * 49 ± 5*† 50 ± 5*†
WRmax, W 291 ± 45 316 ± 49 * 336 ± 51*† 340 ± 48*†
Estimated θ̂L , L/min 1.99 ± 0.26 2.15 ± 0.34 * 2.38 ± 0.46*† 2.35 ± 0.45*†
Estimated θ̂L , W 117 ± 21 133 ± 23 * 158 ± 30*† 158 ± 25*†
Time-to-fatigue, s 209 ± 51 283 ± 38 * 370 ± 78*† 386 ± 63*†
CON
 O 2max , L/min
V 3.95 ± 0.37 ‡ 3.89 ± 0.35 3.89 ± 0.41 3.98 ± 0.38
 O 2max , mLkg min 48 ± 4 ‡
V -1 -1
47 ± 4 48 ± 5 49 ± 4
WRmax, W 319 ± 38 321 ± 44 320 ± 48 318 ± 42
Estimated θ̂L , L/min 2.15 ± 0.21 2.16 ± 0.16 2.16 ± 0.11 2.15 ± 0.11
Estimated θ̂L , W 116 ± 18 122 ± 17 129 ± 23 ‡ 128 ± 15 ‡
Time-to-fatigue, s 207 ± 58 206 ± 29 ‡ 198 ± 38 ‡ 195 ± 32 ‡

937 Values are means ± SD; V  O 2max , maximal O2 uptake; WRmax, work rate at maximal
938 O2 uptake; θ̂L , lactate threshold. PRE, pre-training; MID, mid-training; POST, post-
939 training; VER, verification. *Significant (P < 0.05) difference from PRE. †Significant
940 (P < 0.05) difference from PRE and MID. ‡Significant (P < 0.05) difference from
941 HIT for corresponding testing point.
 44

942 Table 3A. V O 2p kinetics parameters for lower step (LS) and upper step (US) moderate-intensity exercise transitions, at pre-training,
943 mid-training, post-training and verification testing points

Parameter PRE MID POST VER

LS US LS US LS US LS US
HIT
 O 2p bsl L/min
V 1.00 ± 0.09 1.50 ± 0.29 # 0.98 ± 0.09 1.46 ± 0.15 # 0.98 ± 0.10 1.47 ± 0.17 # 0.94 ± 0.06 1.43 ± 0.13 #
 O 2p ss, L/min
V 1.44 ± 0.14 1.96 ± 0.24 # 1.45 ± 0.15 1.97 ± 0.27 # 1.47 ± 0.17 2.00 ± 0.26 # 1.43 ± 0.13 1.94 ± 0.23 #
Amp, L/min 0.44 ± 0.12 0.52 ± 0.11 # 0.47 ± 0.13 0.52 ± 0.13 # 0.49 ± 0.12 0.54 ± 0.11 # 0.49 ± 0.12 0.51 ± 0.11
TD, s 13 ± 4 5 ± 10 # 16 ± 5 * 6±6# 20 ± 2 * 11 ± 6 † # 20 ± 6 * 13 ± 9 † #
τV O 2p , s 24 ± 6 45 ± 5 # 18 ± 3 * 35 ± 8 *# 15 ± 2 * 28 ± 7 †# 14 ± 3 † 25 ± 8 †#
C95, L/min 5±1 5 ±2 4±2 5±2 4±1 4±1 3±1 4±2
ΔV  O 2p ss/ΔWR,
8.4 ± 1.2 10.0 ± 1.0 # 8.9 ± 1.3 9.9 ± 1.3 # 9.3 ± 1.1 10.3 ± 0.4 # 9.5 ± 0.1 9.7 ± 0.8
-1 -1
mLmin W
O2 deficit, mL 265 ± 58 423 ± 125 # 249 ± 89 347 ± 117 # 270 ± 78 329 ± 76 *# 246 ± 73 303 ± 100 †#
CON
 O 2p bl, L/min
V 1.05 ± 0.09 1.49 ± 0.12 # 1.03 ± 0.11 1.46 ± 0.12 # 1.00 ± 0.14 1.43 ± 0.12 # 0.94 ± 0.08 1.38 ± 0.10 #
 O 2p ss, L/min
V 1.49 ± 0.12 2.00 ± 0.19 # 1.46 ± 0.12 1.95 ± 0.17 # 1.43 ± 0.12 1.64 ± 0.63 # 1.38 ± 0.10 1.89 ± 0.17 #
Amp, L/min 0.44 ± 0.08 0.51 ± 0.08 # 0.42 ± 0.08 0.49 ± 0.08 # 0.44 ± 0.09 0.48 ± 0.07 0.44 ± 0.11 0.51 ± 0.08 #
TD, s 17 ± 5 10 ± 6 # 10 ± 14 8±3# 16 ± 8 10 ± 5 # 13 ± 6 4±7#
τV O 2p , s 22 ± 13 38 ± 14 # 24 ± 9 38 ± 13 # 23 ± 10 36 ± 7 #‡ 22 ± 9 40 ± 10 #‡
C95, L/min 5±5 5±2 6±3 6±1 5±1 5±1 6±3 6±2
ΔV  O 2p ss/ΔWR,
8.9 ± 1.0 10.3 ± 0.7 # 8.6 ± 1.3 10.1 ± 1.5 # 8.9 ± 1.3 9.9 ± 1.7 9.0 ± 1.6 10.3 ± 0.8 #
-1 -1
mLmin W
O2 deficit, mL 298 ± 65 382 ± 53 # 221 ± 72 398 ±77 # 285 ± 49 335 ± 97 259 ± 66 371 ± 59 #
 45

944  O 2p are means ± SD; V

Values
V  O 2p    2p O
2p bsl, baseline V O 2p ; V O 2p ss, steady-state V O V ; Amp,
2p amplitude
V O 2p of V O 2p response; TD, time delay; τ
945 , time constant for response; C95, 95% confidence interval of τ ;Δ ss/ΔWR, functional gain. *Significant (P <
946 0.05) difference from PRE. †Significant (P < 0.05) difference from PRE and MID. #Significant (P < 0.05) difference from LS.
947 ‡Significant (P < 0.05) difference from HIT for corresponding testing point.
 46

948 Table 3B. V O 2p kinetics parameters for single step (SS) moderate-intensity exercise
949 transitions, at pre-training and verification testing points
950

Parameter PRE VER

SS SS
HIT
 O 2p bsl L/min
V 1.05 ± 0.13 1.00 ± 0.08
 O 2p ss, L/min
V 2.00 ± 0.20 1.97 ± 0.24
Amp, L/min 0.94 ± 0.19 0.97 ± 0.24
TD, s 9±6 16 ± 3*
τV O 2p , s 32 ± 7 19 ± 3*
C95, L/min 4±1 3±1
ΔV  O 2p ss/ΔWR,
9.1 ± 0.9 9.3 ± 1.0
-1 -1
mLmin W
O2 deficit, mL 661 ± 110 527 ± 102*
CON
 O 2p bl, L/min
V 1.11 ± 0.11 0.98 ± 0.10*
 O 2p ss, L/min
V 2.01 ± 0.12 1.88 ± 0.07*
Amp, L/min 0.99 ± 0.17 0.90 ± 0.11
TD, s 9±6 7 ± 13
τV O 2p , s 34 ± 11 32 ± 13 ‡
C95, L/min 3±1 4±2
ΔV  O 2p ss/ΔWR,
10.0 ± 1.9 9.3 ± 1.6
mLmin-1W-1
O2 deficit, mL 638 ± 108 617 ± 79

951 Values are means ± SD; V O 2p bsl, baseline V O 2p ; V O 2p ss, steady-state V O 2p ; Amp,
952 amplitude of V O 2p response; TD, time delay; τ V O 2p , time constant for V O 2p
953 response; C95, 95% confidence interval of τ V O 2p ; Δ V O 2p ss/ΔWR, functional gain.
954 *Significant (P < 0.05) difference from PRE. ‡Significant (P < 0.05) difference from
955 HIT for corresponding testing point.

956
 47

957 Table 4A. [HHb] kinetics parameters for lower step (LS) and upper step (US) moderate-intensity exercise transitions, at pre-training,
958 mid-training, post-training and verification testing points

959

Parameter PRE MID POST VER

LS US LS US LS US LS US
HIT
[HHb]bsl, AU 28.1 ± 5.1 34.4 ± 8.1 # 27.3 ± 4.6 32.1 ± 6.7 # 30.7 ± 5.2 37.3 ± 6.8 # 31.9 ± 3.5 39.2 ± 7.0 #
[HHb]ss, AU 34.5 ± 7.8 40.1 ± 9.5 # 31.9 ± 6.4 38.0 ± 8.2 # 37.1 ± 6.9 43.8 ± 9.3 # 37.7 ± 6.6 46.4±11.2 #
[HHb]
34.2 ± 7.5 41.6 ± 10.3 # 32.2 ± 6.9 38.8 ± 8.6 # 37.3 ± 6.8 44.2 ± 9.2 # 38.0 ± 5.3 46.8 ± 8.9 #
End-Stp, AU
TD, s 11 ± 3 9±2# 12 ± 2 10 ± 2 12 ± 3 8±2# 13 ± 2 9±3#
τ[HHb], s 8±4 23 ± 12 # 8±5 23 ± 8 # 9±8 20 ± 14 # 8±6 20 ± 13 #
τ', s 18 ± 5 32 ± 11 # 20 ± 4 33 ± 6 # 21 ± 6 28 ± 12 21 ± 5 28 ± 13
∆[HHb]ss
14 ± 9 14 ± 8 11 ± 6 14 ± 9 14 ± 8 14 ± 6 12 ± 5 15 ± 6
 O 2p ss
/Δ V
CON
[HHb]bsl, AU 27.2 ± 6.2 32.4 ± 5.6 # 27.1 ± 3.1 31.9 ± 4.3 # 27.7 ± 6.6 31.0 ± 6.8 # 26.7 ± 7.2 31.3 ± 7.7 #‡
[HHb]ss, AU 32.5 ± 7.3 38.2 ± 8.0 # 32.1 ± 3.6 38.0 ± 7.4 # 31.5 ± 6.9 36.3 ± 10.9 30.9 ± 8.4 36.1±10.5 #
[HHb]
32.4 ± 5.6 39.4 ± 7.9 # 31.9 ± 4.4 38.5 ± 7.5 # 31.0 ± 6.8 36.8 ± 11.0 # 30.7 ± 6.6 36.7 ± 9.6 #
End-Stp, AU
TD, s 11 ± 3 7±2# 11 ± 2 7 ± 3 #‡ 11 ± 4 7±1 13 ± 1 9±1#
τ[HHb], s 9±3 25 ± 10 # 7±2 19 ± 7 # 7±4 19 ± 6 # 9±3 23 ± 3 #
τ', s 20 ± 2 32 ± 9 # 18 ± 2 26 ± 7 18 ± 2 26 ± 5 # 22 ± 3 31 ± 3 #
 48

∆[HHb]ss
13 ± 3 15 ± 7 12 ± 6 15 ± 9 9±4 13 ± 10 12 ± 7 13 ± 7
 O 2p ss
/Δ V

960 Values are means ± SD; [HHb]bsl, baseline; [HHb]ss, fitted steady-state; [HHb] End-Stp, end-transition [HHb]; TD, time delay;
961 τ[HHb], time constant for [HHb] response; τ', effective time constant (τ[HHb] + TD); ∆[HHb]ss, /Δ V O 2p ss, gain. No significant
962 changes occurred over time (P > 0.05). #Significant (P < 0.05) difference from LS. ‡Significant (P < 0.05) difference from HIT for
963 corresponding testing point.
 49

964 Table 4B. [HHb] kinetics parameters for single step (SS) moderate-intensity exercise transitions, at pre-training and verification
965 testing points

966

Parameter PRE VER

SS SS
HIT
[HHb]bsl, AU 27.0 ± 5.2 30.5 ± 2.4
[HHb]ss, AU 39.6 ± 10.3 44.6 ± 7.1
[HHb]
40.4 ± 9.7 45.5 ± 7.3
End-Stp, AU
TD, s 10 ± 2 10 ± 2
τ[HHb], s 10 ± 4 11 ± 6
τ', s 19 ± 5 21 ± 7
∆[HHb]ss
15 ± 9 17 ± 10
 O2ss
/Δ V
CON
[HHb]bsl, AU 26.7 ± 3.8 27.0 ± 8.3
[HHb]ss, AU 38.9 ± 8.4 36.6 ± 12.1
[HHb]
39.9 ± 9.2 35.3 ± 8.5
End-Stp, AU
TD, s 8±3 8±2
τ[HHb], s 12 ± 3 12 ± 3
τ', s 20 ± 2 20 ± 2
∆[HHb]ss
13 ± 4 14 ± 10
 O2ss
/Δ V

967 Values are means ± SD; [HHb]bsl, baseline; [HHb]ss, fitted steady-state; [HHb] End-Stp, end-transition [HHb]; TD, time delay;
968 τ[HHb], time constant for [HHb] response; τ', effective time constant (τ[HHb] + TD); ∆[HHb]ss, /Δ V O 2p ss, gain. No significant
969 changes occurred over time (P > 0.05). No significant differences existed amongst any variable between HIT and CON.

A

Figure 1

A

Figure 2


Figure 3

Lower Step (LS)

PRE

VER

Figure 4A

Upper Step (US)

PRE

VER

Figure 4B

Single Step (SS)

PRE

VER

Figure 5