Introduction

An abrupt increase in work rate (WR) produces an instantaneous increase in adenosine triphosphate (ATP) hydrolysis and thus the requirement for ATP re-synthesis; however, the adjustment of muscle oxidative phosphorylation (i.e., muscle VO2 kinetics) towards the new steady-state requirement is exponential, rather than immediate (Linnarsson 1974; Whipp and Wasserman 1972). The profile of this exponential adjustment at the muscle can be reliably characterized (to within ~10 %) by pulmonary gas-exchange measures (Grassi et al. 1996) and is generally characterized by both its rate (quantified as the phase II pulmonary VO2 time constant; τVO2p) and the VO2p amplitude (i.e., ΔVO2p from baseline to steady-state VO2p) or VO2p functional gain (G = ΔVO2p/ΔWR). Based on early observations of similar τVO2p values (Whipp 1971) in response to different WR transitions within the moderate-intensity domain (i.e., below the estimated lactate threshold; θ L) as well as on-/off-transient symmetry (Griffiths et al. 1986; Paterson and Whipp 1991) in response to a single moderate-intensity exercise transition, it was believed that the characteristics of the VO2p kinetic response were independent of WR within the moderate-intensity domain (Grassi et al. 1996; Linnarsson 1974; Whipp and Wasserman 1972).

The notion that the parameters describing VO2p kinetics within the moderate-intensity domain are “WR independent” has not been thoroughly assessed. The WR independence of the VO2p kinetics parameters has been challenged by the findings of several studies that have used a “double step” protocol in both young (Brittain et al. 2001; Hughson and Morrissey 1982; MacPhee et al. 2005) and older (Spencer et al. 2011a) adults; this “double step” protocol implies arrival at the final WR following two “step” transitions of equal magnitude, rather than a single “step” transition spanning the entire WR change. A feature shared amongst these studies is the observation of a greater τVO2p values (i.e., slower adjustments) and greater functional G when exercise is initiated from an elevated WR, even within the moderate-intensity domain. As such, the primary focus of these studies (as well as that of Bowen et al. (2011)) was to elucidate the underlying cause of this slowing. Importantly, however, a consistent finding that has received somewhat less attention in the aforementioned studies was the principal cause of the trend (though not always significant) for smaller τVO2p values (i.e., faster adjustments) and smaller functional G when the WR increment is smaller (where the pre-transition WR was constant and relatively low (i.e., rest or 20 W)). That is, in each of these studies, a “lower step” (i.e., generally to a WR corresponding to ~45 % of θ L) was compared with a “full step” (i.e., generally with a WR corresponding to ~90 % of θ L) where transitions were initiated from identical (low) baseline WRs. A concern with these results is the particularly small WR increment of the “lower step” and whether the lower step VO2p kinetics parameters are accurate given the low signal-to-noise of the data. For example, both MacPhee et al. (2005) and Spencer et al. (2011a) reported 95 % confidence intervals for estimates of τVO2p (CI95 τVO2p) that were significantly greater for “lower” and “upper” (i.e., small WR increment) steps, compared with a “full step” (i.e., larger WR increment).

In addition to measures of VO2p throughout exercise transitions, near-infrared spectroscopy (NIRS) has been used extensively of late to monitor changes in muscle deoxygenation (Δ[HHb]), which reflects the balance between O2 delivery and O2 utilization (Ferrari et al. 2011). By pairing these respective signals (i.e., Δ[HHb] and VO2p) and considering the rate of their adjustment relative to one another, our laboratory has recently identified an “overshoot” in the Δ[HHb]/VO2 ratio, likely reflecting a transient local muscle O2 delivery to O2 utilization mismatch (Murias et al. 2011d). This Δ[HHb]/VO2 overshoot has been shown to be attenuated or abolished with training (Murias et al. 2010, 2011a) and heavy-intensity ‘priming’ exercise (Murias et al. 2011b; Spencer et al. 2012) and is exacerbated by exposure to acute hypoxia (Spencer et al. 2012).

Whereas Wilkerson et al. (2004) have described the effects of WR on phase II τVO2p and functional G across a broad range of exercise intensity domains (i.e., from 60 % θ L to 120 % of peak VO2), and Scheuermann and Barstow (2003) described the VO2 kinetic response to exercise at (i.e., 100 %) or near (i.e., 90 and 110 %) VO2max, the effects of WR on VO2 kinetic parameters within the moderate-intensity domain have not yet been thoroughly described. Furthermore, it remains unclear whether WR increment might affect the Δ[HHb]/VO2 ratio, which could provide important insights into mechanisms underlying possible differences in the phase II τVO2p and functional G. Thus, the purpose of the present study was to systematically examine the role of WR increment (when initiated from a constant low WR of 20 W to five different moderate-intensity WRs between 50 and 130 W) on both τVO2p and functional G in a group of healthy, young adults. Further, with the hypothesis of both smaller τVO2p and functional G during transitions to lower WRs, we sought to investigate the potential mechanism(s) using measures of local muscle deoxygenation (to assess the balance between O2 delivery and O2 utilization) and to determine whether this mechanism differed between those individuals who presented with fast compared with slower VO2p kinetics.

Methods

Subjects

Fourteen young men (24 ± 5 years; 80 ± 12 kg; 180 ± 6 cm; mean ± SD) volunteered and gave written informed consent to participate in the study. All procedures were approved by The University of Western Ontario Research Ethics Board for Health Sciences Research Involving Human Subjects. All subjects were non-smokers and were physically active. Additionally, no subjects were taking medications that are known to affect the cardiorespiratory, metabolic or haemodynamic responses to exercise.

Protocol

Initially, subjects reported to the laboratory to perform a ramp incremental test (25 W/min) to the limit of tolerance on a cycle ergometer (model: H-300-R Lode; Lode B.V., Groningen, Holland) for determination of peak VO2 (VO2peak) and θ L. θ L was defined as the VO2 at which CO2 output (VCO2) began to increase out of proportion to VO2 with a systematic rise in minute ventilation-to-VO2 ratio and end-tidal PO2, whereas minute ventilation-to-VCO2 ratio and end-tidal PCO2 were stable (Beaver et al. 1986; Whipp et al. 1986). Subsequent to the incremental test, subjects returned to the laboratory on six occasions during which a total of four repetitions of “step” changes in WR were completed per visit. During each visit, subjects completed a pair of successive moderate-intensity leg cycling transitions from a 6-min baseline WR of 20 W to 6 min at a WR corresponding to one of 50, 70, 90, 110 or 130 W separated by 6 min of 20 W cycling; following 20 min of seated recovery, subjects continued on to complete a second pair of identical transitions. Subjects were instructed to maintain a pedal cadence between 60 and 70 revolutions per minute (rpm) throughout 20 W baseline cycling and during each of the moderate-intensity cycling WRs. Cycle cadence was displayed continuously and the investigator ensured pedal cadence was in the 60–70 rpm range. The decision to test subjects at the same absolute WRs, rather than similar relative (i.e., % of θ L) WRs, was made to avoid perpetuating potential errors in the original estimation of θ L (which is likely to have a small, inherent error when estimated from gas-exchange data); that is, if θ L was under- or over-estimated for a given subject, this error would be repeated five times when prescribing individual WRs. For each individual, the step transitions from 20 to 130 W were performed first to verify that this WR was within the moderate-intensity domain (with the absence of a VO2p slow component); after this visit, the order of the WR transitions was randomized. For the 20- to 50-W transitions, a total of eight repetitions (i.e., 2 visits) were performed, as smaller WR transitions have been associated with reduced signal-to-noise ratios and thus (perhaps) less confidence in parameter estimates; for all other WRs, four transitions (i.e., 1 visit) were performed. We have previously shown that the effect of previous moderate-intensity transitions on the VO2p kinetics response of subsequent moderate-intensity transitions is negligible (Spencer et al. 2011b). Each visit to the laboratory was separated by at least 24 h.

Measurements

Gas-exchange measurements were similar to those previously described (Babcock et al. 1994). Briefly, inspired and expired flow rates were measured using a low dead space (90 mL) bidirectional turbine (Alpha Technologies VMM 110) which was calibrated before each test using a syringe of known volume. Inspired and expired gases were continuously sampled (50 Hz) at the mouth and analyzed for concentrations of O2, CO2 and N2 by mass spectrometry (Innovision, AMIS 2000, Lindvedvej, Denmark) after calibration with precision-analyzed gas mixtures. Changes in gas concentrations were aligned with gas volumes by measuring the time delay for a square-wave bolus of gas passing the turbine to the resulting changes in fractional gas concentrations as measured by the mass spectrometer. Data were transferred to a computer, which aligned concentrations with volume information to build a profile of each breath. Breath-by-breath alveolar gas exchange was calculated using algorithms of Beaver et al. (1981).

Local muscle deoxygenation profiles of the quadriceps vastus lateralis muscle were made with continuous-wave NIRS (Hamamatsu NIRO 300, Hamamatsu Photonics, Hamamatsu, Japan). The physical principles of tissue spectroscopy and the manner in which these are applied have been explained by DeLorey et al. (2003). Briefly, optodes were placed on the belly of the muscle midway between the lateral epicondyle and greater trochanter of the right femur. The system consisted of both an emission probe that carries NIR light from the laser diodes and detector probe (interoptode spacing 5 cm); optodes were housed in an optically dense plastic holder and secured on the skin surface with tape and then covered with an optically dense, black vinyl sheet, thus minimizing the intrusion of extraneous light. The thigh was wrapped with an elastic bandage to minimize movement of the optodes. Four laser diodes (λ = 775, 810, 850, and 910 nm) were pulsed in a rapid succession and the light returning from the tissue was detected by the photodiode for online estimation and display of the concentration changes from the resting baseline for oxyhaemoglobin (Δ[HbO2]), deoxyhaemoglobin (Δ[HHb]) and total haemoglobin (Δ[Hbtot]). One limitation of the NIRS technique is the overlapping absorption spectra of muscle myoglobin (Mb) and Hb (Bowen 1949); this makes separation of these absorbers difficult. However, the contribution of Mb to light absorption changes is estimated to be ~10 %, which lends to the interpretation that light absorption changes with NIRS are attributable mainly to the oxygenation status of Hb (Kalliokoski et al. 2006). Changes in light intensities were recorded continuously at 2 Hz and transferred to a computer for later analysis. The NIRS-derived signal was zero set with the subject sitting in a resting steady-state on the cycle ergometer prior to the onset of baseline exercise and changes in the concentration are reported as a delta (∆) in arbitrary units (a.u.).

Data analysis

VO2p data were filtered by removing aberrant data points that lay outside 4 SD of the local mean; the justification for this filtering process was provided by Lamarra et al. (1987), who demonstrated that “noise” observed within the VO2p signal conformed to a predictable Gaussian distribution, independent of WR. The data for each transition were linearly interpolated to 1 s intervals and time-aligned such that time zero represented the onset of exercise. Data from all same-WR transitions were ensemble-averaged to yield five averaged responses for each subject (i.e., one for each WR). These transitions were further time-averaged into 5 s bins to provide five time-averaged responses for each subject. (Ensemble- and time-averaged responses for 20 to 50 W transitions using only 4 and only 6 transitions were also generated in addition to that generated using all 8 transitions.) Baseline VO2p (VO2pBSLN) was calculated as the average VO2p collected in the 2 min before an increase in WR. The on-transient responses for VO2p were modelled using the following equation:

$$ Y_{(t)} = Y_{\text{B}} + A \, (1 - {\text{e}}^{{ - (t - {\text{TD}})/\tau }} ) $$
(1)

where Y (t) represents the VO2p for any given time; Y B is the VO2p at baseline; A is the amplitude of the VO2p response; t is a given amount of time; τ represents the time required to attain 63 % of the steady-state amplitude; and TD represents the time delay.

After excluding the initial 20 s of data (which, while not necessarily reflecting the exact duration of the ‘cardiodynamic phase’ in each individual, is most likely to avoid inclusion of data points from phase I VO2p in the fitting of phase II VO2p (Murias et al. 2011c), while still allowing TD to vary freely (in order to optimize accuracy of parameter estimates), VO2p data were modelled from 20 s to 4 min (240 s) of the step-transition; this ensured that each subject had attained a VO2p steady-state (VO2pSS), yet did not bias the model fit during the on-transient (Bell et al. 2001; Murias et al. 2011c). The model parameters were estimated by least-squares nonlinear regression (Origin, OriginLab Corp., Northampton, MA, USA) in which the best fit was defined by minimization of the residual sum of squares and minimal variation of residuals around the Y-axis (Y = 0). The 95 % confidence interval for the estimated time constant (CI95 τVO2p) was determined after preliminary fit of the data with Y B, A and TD constrained to the best-fit values and the τ allowed to vary.

In order to investigate whether the changes in τVO2p and functional G differed between those individuals who presented with fast compared with slower VO2p kinetics, the sample was sub-divided into two groups. A “cut-off” of τVO2p = 25 s (during transitions to 130 W) was selected to separate the two groups. This value was selected in an effort to differentiate those individuals in whom an O2 delivery-dependent limitation of τVO2p was likely compared with those individuals in whom an O2 delivery-dependant limitation of τVO2p was less likely based on recent findings from our laboratory (Murias et al. 2010, 2011b, d; Spencer et al. 2012). While this value is not completely congruent with our findings of an “O2 delivery independent” limitation of τVO2p when τVO2p ≤ 21 s (Murias et al. 2011d), strict adherence to this 21 s value would have created a Fast group with only four subjects (and 10 in the Slower); this jeopardized the statistical power of the present study. Further, a strong linear relationship (r = 0.91) was established between the normalized ∆[HHb]/VO2 ratio and τVO2p (Murias et al. 2011d), suggesting that in those individuals with τVO2p values between 21 and 25 s, any O2 delivery to O2 utilization mismatch during the exercise on-transient would be modest.

Five NIRS-derived ∆[HHb] responses (i.e., one for each WR) were generated for each subject. The Δ[HHb] profile has been described to consist of a time delay at the onset of exercise, followed by an increase in the signal with an “exponential-like” time-course (DeLorey et al. 2003). The time delay for the ∆[HHb] response (TD Δ[HHb]) was determined using second-by-second data and corresponded to the time, after the onset of exercise, at which the ∆[HHb] signal began a systematic increase from its nadir value. Determination of the TD ∆[HHb] was made on individual trials and averaged to yield a value for each individual in each of the five “conditions.” The ∆[HHb] data were modelled from the end of the TD ∆[HHb] to 90 s of the transition using an exponential model as described in Eq. (1). As previously described by duManoir et al. (2010), different fitting strategies (i.e., 90–180 s) resulted in minimal differences (<2 s) in estimates of the Δ[HHb] time constant (τ∆[HHb]) and the early exponential increase in ∆[HHb] was well characterized in the 90 s following exercise onset. The τ∆[HHb] described the time course for the increase in ∆[HHb], while the overall change of the effective ∆[HHb] (τ′∆[HHb] = TD ∆[HHb] + τ∆[HHb]) described the overall time course of the ∆[HHb] from the onset of exercise.

Additionally, the second-by-second ∆[HHb] and VO2p data were normalized for each subject (0 %, representing the 20-W baseline value, and 100 %, representing the post-transition steady-state of the response). The normalized VO2p was left-shifted by 20 s to account for the approximate phase I–phase II transition so that the onset of exercise coincided with the beginning of phase II VO2p (Murias et al. 2011d), which has been previously described to coincide with muscle VO2 within 10 % (Grassi et al. 1996). Following normalization and time-alignment, data were further averaged into 5 s bins for statistical comparison of the rate of adjustment for ∆[HHb] and VO2p. Additionally, an overall ∆[HHb]/VO2 ratio for the adjustment during the exercise on-transient was derived for each individual as the average value from 20 to 120 s into the transition. The start point was selected to be 20 s to begin the analysis at the time region when the ∆[HHb] and VO2p signals meet, reflecting the TD ∆[HHb] and early adjustment of ∆[HHb]. An end point of 120 s was selected as the time point at which the ∆[HHb]/VO2 ratio had reached a steady-state value of 1.0.

Statistics

Data are presented as mean ± SD. Independent samples t tests were used to detect between group differences in subject characteristics. Two-way (WR and group) repeated measures analyses of variance (ANOVA) were used to determine statistical significance for the dependent variables (i.e., VO2p and Δ[HHb] kinetic parameter estimates). Tukey’s post hoc analysis was used when significant differences were detected for the main effects of each dependent variable. Determination of whether a Δ[HHb]/VO2 ratio overshoot was significant was based upon a comparison (t test) of the WR- or WR by group-specific mean with a value of 1.0 (with no associated error). Pearson’s product–moment correlation coefficients were used to quantify the strength of relationships between variables. All statistical analyses were performed using SPSS Version 19.0, (SPSS Inc., Chicago, IL, USA). Statistical significance was declared when p < 0.05.

Results

Subject characteristics are presented in Table 1. Peak WR was 361 ± 37 W, which yielded a VO2peak of 4.13 ± 0.40 L min−1 (or 52.3 ± 6.8 mL kg−1 min−1). When the sample (n = 14) was sub-divided into those with Fast (τVO2p at 130 W <25 s; n = 6; VO2peak = 56.7 ± 4.8 mL kg−1 min−1) and Slower (τVO2p at 130 W >25 s; n = 8; VO2peak = 49.1 ± 6.4 mL kg−1 min−1) VO2p kinetics, a between-group difference (p < 0.05) was observed for ‘relative’ VO2peak, but no other characteristics. The VO2p responses elicited by transitions to 50, 70, 90, 110 and 130 W corresponded to 49 ± 6, 57 ± 7, 66 ± 7, 75 ± 7 and 85 ± 8 % of θ L. Importantly, no between-group differences were observed when WR was expressed as a relative % of θ L (p = 0.43), rather than in absolute terms (i.e., W).

Table 1 Subject characteristics
Full size table

Figure 1 displays the profiles of VO2p (5 s averaged, group mean data) for the Fast and Slower groups for exercise transitions from 20 to 50, 70, 90, 110 and 130 W. Group mean (±SD) VO2p kinetic parameter estimates (i.e., derived from the mono-exponential non-linear regressions) for each of the five constant load WRs are illustrated in Fig. 2. VO2pBSLN was consistent both amongst WRs and between groups (p > 0.05). Whereas no between-group differences were observed in the VO2p amplitude (VO2pAMP) response, there was a significant main effect of WR (p < 0.05) such that each progressive increase in WR yielded a greater VO2pAMP response. A consequence of invariant VO2pBSLN values, along with progressively increasing VO2pAMP values with increasing WRs, was a progressively increasing VO2pSS response (p < 0.05; not depicted in Fig. 2) which was observed in both the Fast and Slower groups. Transitions to greater WRs were associated with decreased CI95 τVO2p (p < 0.05). There was a significant increase (p < 0.05) in functional G during transitions to increasing WRs within the moderate-intensity domain; this finding was not associated with any between-group differences (p > 0.05). When considered as a single group, there was no main effect of WR on τVO2p; however, both a main effect for group (i.e., Fast vs. Slower, p < 0.05) and a WR by group interaction (p < 0.05) were observed. Post hoc comparisons revealed significantly greater τVO2p during transitions to 110 and 130 W (by design) in the Slower group, but not for transitions to 50, 70 and 90 W; these findings suggest a trend for increasing τVO2p values during transitions to a higher WR in the Slower group which was absent in the Fast group (or decreasing τVO2p values during transitions to a higher WR in the Fast group which was absent in the Slower group).

Fig. 1
figure 1

Group mean (5 s averaged data) VO2p profiles for transitions from 20 to 50 W (top), 70, 90, 110 and 130 W (bottom) in Fast (left) and Slower (right) groups. Mono-exponential models from group mean data superimposed. Exercise onset denoted by solid vertical line. Dashed horizontal line indicates predicted VO2p assuming functional G of 10 mL min−1 W−1

Full size image
Fig. 2
figure 2

Mean (±SD) VO2p kinetic parameter estimates for transitions from 20 to 50, 70, 90, 110 and 130 W in All subjects (filled circle). *p < 0.05 from 50 W; # p < 0.05 from 70 W; † p < 0.05 from 90 W; ‡ p < 0.05 from 110 W. Fast (open triangle) and Slower (open square) groups are depicted when main effect of GROUP ($ p < 0.05) or WR*GROUP interaction (§ p < 0.05) were detected

Full size image

The effects of transition WR magnitude on Δ[HHb] kinetic parameter estimates are displayed in Fig. 3. The baseline Δ[HHb] (Δ[HHb]BSLN) was not different for any WR, but when the two groups were considered separately, the Δ[HHb]BSLN was lower (p < 0.05) in the Fast compared with the Slower group; the between group difference observed (p < 0.05) is of little consequence, since the (Δ) units are arbitrary. Greater WR transitions were associated with lower 95 % confidence intervals (reflecting greater confidence in parameter estimation) for the τΔ[HHb] (CI95 τΔ[HHb]; p < 0.05), as well as greater (p < 0.05) Δ[HHb] amplitudes (Δ[HHb]AMP) and steady-state Δ[HHb] (Δ[HHb]SS) responses; again, the between-group differences (Δ[HHb]AMP and Δ[HHb]SS; p < 0.05) and WR*group interaction (Δ[HHb]AMP; p < 0.05) are inconsequential. Modest but significant reductions in both τΔ[HHb] and TD Δ[HHb] were observed with increasing WR transitions (p < 0.05), and as a result, τ′Δ[HHb] decreased with increasing WR transitions (p < 0.05). Whereas a main effect of group was observed for τΔ[HHb] (with Slower having a greater τΔ[HHb] response), this between-group difference was absent in both the TD Δ[HHb] and τ′Δ[HHb] responses.

Fig. 3
figure 3

Mean (±SD) Δ[HHb] kinetic parameter estimates for transitions from 20 to 50, 70, 90, 110 and 130 W in All subjects (filled circle). *p < 0.05 from 50 W; # p < 0.05 from 70 W; † p < 0.05 from 90 W; ‡ p < 0.05 from 110 W. Fast (open triangle) and Slower (open square) groups are depicted when main effect of GROUP ($ p < 0.05) or WR*GROUP interaction (§ p < 0.05) were detected

Full size image

No effect of WR was observed with respect to the Δ[HHb]/VO2 ratio (p > 0.05); a main effect of group was identified (p < 0.05) such that the Slower group had a greater Δ[HHb]/VO2 ratio than the Fast group (Fig. 4). Furthermore, the Δ[HHb]/VO2 ratio observed in the Fast group did not differ (p > 0.05) from 1.0 during transitions to any WR. Similar to previous studies from our laboratory (Murias et al. 2011b, d), the Δ[HHb]/VO2 ratio was associated with τVO2p in the present study (r = 0.71; p < 0.05; Fig. 5).

Fig. 4
figure 4

Group mean (5 s averaged data) Δ[HHb]/VO2 profiles for transitions from 20 to 50 W (top), 70, 90, 110 and 130 W (bottom) in Fast (left) and Slower (right) groups. Dashed horizontal line indicates Δ[HHb]/VO2 = 1.0

Full size image
Fig. 5
figure 5

Relationship (r = 0.71) between the ∆[HHb]/VO2 ratio and τVO2p for all 14 subjects during transitions from 20 to 50, 70, 90, 110 and 130 W. ∆[HHb]/VO2 ratio was derived for each individual as the average of the 21 5 s ratio values from 20 to 120 s

Full size image

Table 2 illustrates the effects of performing “additional” repetitions of small WR transitions (e.g., 20 to 50 W) on VO2p kinetic parameter estimates. Briefly, no differences (p > 0.05) in any of VO2pBSLN, VO2pAMP, VO2pSS, functional G and τVO2p were observed when parameters were derived from the average of four, six or eight repetitions. Performing either six or eight repetitions yielded significant reductions in the CI95 τVO2p compared with performing only four (p < 0.05; p = 0.13 between 6 and 8 repetitions). Indeed, as shown in Fig. 1, even with eight repetitions the confidence in the τVO2p, (and τΔ[HHb] as shown in Fig. 2) is two- to threefold less for the 50 W WR versus WRs above 100 W.

Table 2 VO2p kinetic parameter estimates for 20 to 50 W transitions when four, six or eight transitions were ensemble-averaged
Full size table

Discussion

The present study sought to systematically examine the role of WR increment (when initiated from a constant low WR of 20 W to WRs in the moderate-intensity exercise domain) on both τVO2p and functional G in a group of healthy, young adults. Further, with the hypothesis of both smaller τVO2p and functional G during transitions to lower WRs, we sought to investigate the potential mechanism and to determine whether this mechanism differed between those individuals who presented with fast compared with slower VO2p kinetics. The main findings of the study were that (1) during transitions to different WRs within the moderate-intensity domain, no differences in τVO2p were observed in this group of subjects (τVO2p ≈ 27 s); (2) when the sample was sub-divided into two groups (i.e., those with fast and slower VO2 kinetics), there was an interaction between WR increment and group such that the τVO2p responses appeared divergent between the Fast and Slower groups, with only the Slower group showing evidence of conformity to WR-dependent alterations in τVO2p previously reported in the literature (i.e., in studies comparing a “lower step” to a “full step”); (3) the Δ[HHb]/VO2 ratio was smaller in the Fast compared with Slower group (suggesting that the primary determinant of τVO2p differs between groups), but was not affected by WR (suggesting that the divergent τVO2p responses between Fast and Slower groups may not have been the result of O2 availability during the exercise on-transient); (4) functional G was progressively greater during transitions of increasing WR increment; this was true for (and similar between) both the Fast and Slower groups, possibly suggesting that the mechanism(s) controlling the magnitude and the rate of the VO2p response to a given WR transition may be dissociated; and (5) performing either six or eight compared with four repetitions of small (i.e., 20 to 50 W) WR transitions improved the confidence in the estimation of τVO2p, yet, there were no changes in other key VO2p kinetic parameter estimates.

The present study was designed to elicit like VO2pBSLN values amongst the different (WR) conditions, as well as progressively greater VO2pAMP and VO2pSS values during transitions to greater WRs; as such, these findings reported in Fig. 1 were to be expected. Thus, this discussion attempts to sort out the interrelations among three important findings of the present study, namely (1) divergent τVO2p responses between Fast and Slower groups with respect to WR; (2) increasing functional G with increasing WR increments (irrespective of group); and (3) group differences in the Δ[HHb]/VO2 ratio, but with no influence of WR. As mentioned, previous studies examining the effects of either WR increment or pre-transition WR have focused much of their attention on discussing the slower adjustment and greater functional G observed when work is initiated from either an elevated WR (Brittain et al. 2001; Hughson and Morrissey 1982; MacPhee et al. 2005; Spencer et al. 2011a) or an elevated metabolic rate (Bowen et al. 2011); nevertheless, these discussions have identified three potential causes for these trends which will form the basis for the present discussion:

  1. 1.

    an insufficient O2 delivery, leading to a slowing of the VO2 response during the exercise on-transient, was favoured by Hughson and Morrissey (1982) and later supported by MacPhee et al. (2005), but was largely refuted by the findings of Spencer et al. (2011a);

  2. 2.

    a hierarchical recruitment pattern favouring recruitment of the fastest kinetic, most efficient fibres to perform small WR transitions, leaving only those inherently slower kinetic, less efficient fibres to address the demands of a subsequent increase in WR was proposed by Brittain et al. (2001); such a system would allow for an ‘intermediate’ rate of adjustment of VO2 and O2 cost (per unit increase in WR) during larger WR transitions;

  3. 3.

    the influence of a potentially less favourable energetic status (i.e., less negative changes in Gibb’s free energy; ΔG ATP) resulting from either an elevated metabolic rate per se (i.e., irrespective of initial WR) (Glancy et al. 2008; Kemp 2008) or the fact that ΔG ATP becomes progressively less negative throughout the transient (as [ADP] and [Pi] rise and [PCr] fall dynamically) and therefore demands an ATP turnover that continues to rise until the steady state is reached was favoured by Bowen et al. (2011).

Considering that smaller τVO2p values often are reported during “half-steps” compared with “full-step” WR transitions within the moderate-intensity domain (when performed from a constant, low baseline WR) (Bowen et al. 2011; Brittain et al. 2001; Hughson and Morrissey 1982; MacPhee et al. 2005; Spencer et al. 2011a), the finding of invariant τVO2p values in response to different moderate-intensity WRs in the present study was somewhat unexpected. However, when the present sample was sub-divided to create Fast and Slower VO2 kinetics groups, their respective responses (with respect to increasing WR increment) were divergent, where the Slower, but not the Fast group tended to conform to the trend observed by previous studies (i.e., with between-group differences noted only at the higher WRs; see Fig. 1). Indeed, in the studies of Brittain et al. (2001) and Spencer et al. (2011a), where ‘slow’ VO2 kinetics (τVO2p >30 s) were reported during large WR transitions within the moderate-intensity domain (90 % of θ L), significant reductions of τVO2p were reported during smaller WR transitions. In the study of MacPhee et al. (2005), in which the group mean τVO2p in response to a large moderate-intensity WR increment was 28 ± 2 s (i.e., somewhat fast relative to those mentioned immediately above), a trend for faster adjustment (to τVO2p = 24 ± 3 s) during smaller WR increment transitions may be inferred, but this trend did not reach the level of statistical significance. In contrast, in a group with relatively fast VO2 kinetics, Wilkerson et al. (2004) reported no differences in τVO2p values derived from transitions to 60 % θ L (23.2 ± 2.1 s) compared with 90 % θ L (21.8 ± 2.3 s). Thus, in agreement with previous findings (albeit from different studies), the divergent responses between Fast and Slower groups suggest that if or when the adjustment of VO2p is already somewhat fast, there may be a reduced potential for further speeding (with smaller WR increments). The subject group of Bowen et al. (2011) might be characterized as “intermediate”, neither fast nor slow (τVO2p = 26 s), but did show a significantly reduced τVO2p (to 20 s) when the WR increment was small compared with large.

Dynamic changes in near-infrared spectroscopy (NIRS)-derived muscle deoxygenation (Δ[HHb]) have been used extensively to provide insights into the balance between local muscle O2 availability and O2 utilization during exercise (Ferrari et al. 2011). During WR transitions in which Δ[HHb] adjusts more rapidly than VO2, what results is a transient period characterized by an increased (relative) reliance on O2 extraction to support a given metabolic rate; this temporary dissociation between the adjustments of Δ[HHb] and VO2 suggests a transient O2 delivery insufficiency. Thus, recent studies from our laboratory (Murias et al. 2010, 2011b, d; Spencer et al. 2012) have used the normalized Δ[HHb]/VO2 ratio as an index of the matching (and mismatching) of O2 delivery to O2 utilization during the exercise on-transient. Cumulatively, these studies suggest that when τVO2p < ~20 s, increases in τVO2p are the result of progressively more severe O2 delivery limitations. In this light, it is not surprising that the Δ[HHb]/VO2 ratio was significantly different between the Fast and Slower groups. The present data do not support a role for an O2 delivery-dependence limitation of τVO2p in the Fast group (where the Δ[HHb]/VO2 ratio is not significantly different from 1.0, implying no “mismatch” of O2 delivery to O2 utilization), but do support such a role in the Slower group (where the Δ[HHb]/VO2 ratio is significantly greater than 1.0, implying an appreciable “mismatch” between O2 delivery to O2 utilization); these findings are in agreement with our previous studies on the topic.

In the present study, the adjustment of Δ[HHb] (i.e., a reduction of both τΔ[HHb] and τ′Δ[HHb]) became progressively faster with increasing WR increments, and importantly, became faster than τVO2p for a given WR. As a result, we expected to observe a significantly greater mismatch between O2 delivery and O2 utilization (as reflected with a greater Δ[HHb]/VO2 ratio) as WR increment increased. The small changes in τVO2p and “noise” in the Δ[HHb]/VO2 ratio may make the index too insensitive to detect differences. Indeed, across all WRs, the “overall” relationship between the Δ[HHb]/VO2 ratio and τVO2p was robust (r = 0.71) and even stronger at the higher WRs (110 W, r = 0.88; 130 W, r = 0.91). Nevertheless, the present data cannot be taken as supporting a role for O2 delivery as an explanation for the divergent τVO2p responses observed in the Fast and Slower groups; that is, implied increases in τVO2p with increasing WR increment (in the Slower group) are likely not related to increasingly insufficient O2 delivery. The present data do not allow us to speculate on the possible roles of fibre recruitment pattern or a lower free energy release with ATP hydrolysis (i.e., less negative ΔG ATP) with increasing WR.

In the present study, increasing WR increments were associated with increasing functional G; interestingly, whereas the τVO2p responses in the Fast and Slower groups were different (i.e., greater for the Slower group) and increasingly divergent at higher WRs (i.e., WR by group interaction observed), the functional G showed no such influence of group. When considered in light of the group—but not WR-mediated differences in the Δ[HHb]/VO2 ratio, it seems clear that local muscle O2 delivery to O2 utilization dynamics do not explain the altered functional G in response to differing WRs. Again, this leaves either an orderly, hierarchical recruitment pattern or the fact that ΔG ATP becomes progressively less negative throughout the transient as possible explanations for the WR-mediated differences in functional G. Taken alone, it is difficult to discern which of one or both of these hypotheses is likely to underlie the WR-mediated differences in functional G; however, neither hypothesis can presently accommodate both invariant τVO2p and simultaneous changes in functional G during transitions to different moderate-intensity WRs as observed when the present sample was considered as a whole. The present data suggest, perhaps for the first time, that factors which determine τVO2p may be independent from factors which determine the functional G during moderate-intensity WR transitions of different magnitudes.

One potential problem in evaluating the role of WR transitions on the kinetics parameters is that smaller WR transitions are likely to have reduced signal-to-noise ratios, and thus, potentially reduced confidence in parameter estimates. Indeed, the present study confirmed the idea that transitions to progressively greater WRs (i.e., those with a greater signal-to-noise ratio (Lamarra et al. 1987) were associated with improved confidence in estimates of τVO2p (and τΔ[HHb], as reflected by the reduced CI95 (i.e., greater confidence in the estimation of the time constant) for both τVO2p and τΔ[HHb]). To this end, Lamarra et al. (1987) established that the effect of interbreath fluctuations (i.e., “noise”) could be dampened by averaging multiple transitions together when characterizing a signal with an inherently low sampling frequency (i.e., breath-by-breath gas exchange). Given that during small WR transitions, the VO2p “signal” is essentially fixed, the only available strategy for improving confidence in kinetic parameter estimates is to have subjects perform multiple repetitions to reduce the “noise”; however, to date, the precise number of transitions required for accurate characterization of small WR transitions has not been described. This information is potentially important when studying individuals in whom small amplitudes are particularly likely, such as heart failure patients (Bowen et al. 2012; Kemps et al. 2010) or those with COPD (Puente-Maestu et al. 2001). The present data illustrate that the inclusion of either two or four “additional” repetitions of 20 to 50 W transitions does not significantly affect any of the key VO2p kinetic parameters; yet, the inclusion of two “additional” (to a total of 6) repetitions significantly reduced the CI95 τVO2p, with no further reduction (p = 0.13) by ensemble-averaging eight transitions. These findings add to those previously reported by our laboratory, showing that during transitions of ~100 W (i.e., 20 to 120 W), the day-to-day reproducibility of τVO2p was not improved by averaging more than three repetitions (Spencer et al. 2011b).

In conclusion, this study presented novel data demonstrating a non-uniform effect of moderate-intensity WR increment on τVO2p, depending upon whether the entire group response was considered or if the divergent responses for individuals with Fast (τVO2 <25 s) and Slower (τVO2 >25 s) VO2 kinetics were considered. Furthermore, this study demonstrated that increasing moderate-intensity WR increments elicits an increased functional G, regardless of the τVO2p response. Neither the divergent τVO2p responses nor the functional G appear to be related to the dynamic matching of O2 delivery to O2 utilization during the exercise on-transient; yet, the observation of invariant τVO2p values with simultaneously increasing functional G in the overall (i.e., not group specific) response suggests that these features of the VO2 response during moderate-intensity exercise may be controlled by separate mechanisms.