Subjects
Among sixty-seven subjects tested (sedentary subjects, recreational and national level cyclists volunteered for this study), fifty-five individuals matching the criterions of one of our groups were selected to participate in this study. These subjects were assigned into six groups (three aerobic fitness levels X two exercise intensities) based on their aerobic fitness level (VO₂max) and exercise intensity performed (maximal intensity - Max or moderate intensity - Mod). The different levels of aerobic fitness were named as follow: Low aerobic fitness level (LF), intermediate aerobic fitness level (IF) and high aerobic fitness level (HF). The number of individuals in each of the 6 groups (N) is described in Table 1. All subjects gave informed consent and the protocol was approved by the university's ethics committee. The subjects were asked to avoid exercise training during the last 2 days before each test and to report to the laboratory at least 3 h after the last meal. Table 1 shows the subject's characteristics of each group.

Experimental design
The subjects performed 1) an incremental test to determine the maximal oxygen uptake (VO₂max) and the intensity associated with the achievement of VO₂max (iVO₂max); 2) constant work-rate cycling exercises to exhaustion at the intensity corresponding to 50% (moderate-intensity exercise - Mod) or 100% (maximal-intensity exercise - Max) of VO₂max to determine the time constant of oxygen uptake response (MRT - mean response time). Thereafter, those individuals matching the VO₂max range values of one of the three aerobic fitness level groups (LF - below 40 ml·kg·min^-1; IF - between 43 and 60 ml·kg·min^-1; HF - above 63 ml·kg·min^-1) were selected. All tests were performed at the same time of day in a climate-controlled (21-22^oC) laboratory. Subjects performed only one test on any given day and the tests were at least 48 h apart from each other but completed within the period of 2 weeks.

Procedures
The tests were conducted on a mechanically braked cycle ergometer (Monark 828E, Stockholm, Sweden) with the pedal frequency maintained constant at 70 rpm (Ciclocomputer Cateye, Japan). Throughout the tests, gas-exchange variables were measured breath-by-breath using a portable gas analyzer system (Cosmed K4b², Rome, Italy). These analyzers have previously been validated over a wide range of exercise intensities (Mclaughlin et al., 2001). Before each test, the O₂ and CO₂ analysis systems were calibrated using ambient air and a gas of known O₂ and CO₂ concentration according to the manufacturer's instructions, while the K4b² turbine flowmeter was calibrated using a 3-l syringe (Cosmed K4b², Rome, Italy). Earlobe capillary blood samples (25 l) were collected into a glass tube and were analyzed for lactate concentration using an automated analyzer (YSI 2300, Ohio, USA).

Subjects performed incremental exercise to volitional exhaustion to determine VO₂max and iVO₂max at the cycle ergometer. The test started at 70 W with increases in power output of 35 W every 3 min. Each subject was encouraged to give a maximum effort and the blood was sampled at the point of exhaustion. Breath-by-breath VO₂ was smoothed using a five-step average filter, and then reduced to 15 s stationary averages for incremental test (Data Management Software, Cosmed, Rome, Italy) to reduce the noise. The VO₂max was defined as the highest 15 s VO₂ value reached during the incremental test. All subjects fulfilled at least two of the following three criteria for VO₂max: 1) respiratory exchange ratio (R) greater than 1.1, 2) a blood lactate concentration greater than 8 mM, and 3) peak HR at least equal to 90% of the age-predicted maximal (Taylor et al., 1955). The iVO₂max was defined as the minimal power output at which VO₂max occurred (Billat and Koralsztein, 1996).

The subjects subsequently performed a constant work rate test (at 50% VO₂max - Mod exercise or at 100% VO₂max - Max exercise) to determine VO₂ kinetics parameters. The subjects were instructed to perform the required intensity until the exhaustion (cadence < 65 rpm for 10 consecutive sec or voluntary exhaustion) at Max exercise or for 6 min at Mod exercise. At the start of cycling exercise, the subjects cycled against zero resistance, until a pedal cadence of 70 rpm was reached, at that point, the pre-selected work rate was imposed and timing began. Oxygen consumption data from constant work-rate tests were filtered for aberrant values (values not included in the interval of average ± 4 SD were excluded).

VO₂ kinetics curve fitting
The VO₂ kinetics during constant work-rate was modeled with a mono-exponential function by an iterative non-linear regression process from the Microcal Origin 6.0 (Northampton, MA, USA) using the following equation (Lamarra et al., 1987):

Where: VO₂(t) represents oxygen uptake at time (t), VO_2b is the baseline value of VO₂ at rest, A is the amplitude of the increase in VO₂ above the baseline value and τ is the mean response time - MRT (defined as the time required to attain 63% of A). In order to increase the confidence of estimated parameters of the equation (Lamarra et al., 1987), VO₂b and A were constrained at the fitting window, respectively, as the mean value at the minute before constant work-rate exercise onset and as the difference between VO₂b and the end exercise VO₂ value.

Statistical analysis
Data are presented as mean + SD. The effects of aerobic fitness on VO₂ kinetics were tested using one-way ANOVA, with Tukey's post-hoc tests where appropriate. ANOVA two-way was applied to compare all other variables across groups (exercise intensity vs. fitness level). Pearson product moment coefficients were used to assess the significance of relationships between VO₂ kinetics and VO₂max. Statistical significance was set at p ? 0.05.

Table 2 contains the indexes of aerobic fitness level obtained from progressive intensity exercise (VO₂max, iVO₂max relative to body mass and absolute iVO₂max) and the mean response time (MRT) that describes VO₂ kinetics during Mod and Max exercises. All indexes of aerobic fitness were significantly different (p < 0.05) between the three levels of aerobic fitness. MRT of the IF and HF groups were significantly faster than MRT of LF group, however, there was no difference between IF and HF groups for this variable (p > 0.05).

Figure 1 and 2 show the coefficient of correlation between MRT of, respectively, Mod (r = -0.51; p < 0.05) and Max (r = - 0.63; p < 0.05) exercises and VO₂max. Although these coefficients were statistically significant, VO₂max explains little of MRT variability at both exercise intensities (Mod, R2 = 26%; Max, R² = 40%).

Figure 3 and 4 show representative VO₂ data and its curve fitting of each fitness levels, respectively, at moderate and maximal exercise. The individual MRT and VO₂max values of these subjects, whose data are shown, are placed on each figure panel.

To our knowledge, this is the first study evaluating the effects of aerobic fitness level on VO₂ kinetics of subjects in the high-end aerobic power range (VO₂max). The faster MRT along with the increase in aerobic fitness level from low to intermediate is in agreement with several reports describing such adaptation in the low-end range of aerobic fitness level (Fukuoka et al., 2002; Phillips et al., 1995; Yoshida et al., 1992). Our main results showing unchanged MRT of Mod and Max exercise as the fitness level increases from intermediate to high are somehow in line with previous short-term longitudinal data obtained from less fit subjects (Carter et al., 2000; Phillips et al,, 1995). Our current findings add to previous data by describing that exercise adaptation leading individuals to the high-end of aerobic fitness level range (VO₂max near of above 65 ml·kg·min^-1) is not able to further improve VO₂ kinetics during both, Mod and Max intensity exercise.

The subjects included in this study were assigned to one of the three groups based on aerobic fitness level (VO₂max). Regarding our purpose, the composition of the high fit group is especially critical, since it must represent the high-end of aerobic fitness level. The indexes often used to describe aerobic fitness and exercise performance (VO₂max and iVO₂max) (Table 2) seem to match the level typical of elite athletes (Caputo and Denadai, 2004; Jeukendrup et al., 2000).

From a time course stand-point, improvements in sub-maximal VO₂ kinetics of low fit individuals seem to occur very early (within a week) after exercise training program onset, even anticipating changes in other indexes of aerobic fitness such as VO₂max and lactate threshold (Phillips et al., 1995). Although this time course response for maximal exercise VO₂ kinetics is still unknown, it seems that as the exercise intensity increases (intensity of exercise test), the effects of aerobic training on these parameters are more likely to occur (Carter et al., 2000; Krustrup et al., 2004). Collectively, these studies cited above show a good frame of the initial response of VO₂ kinetics, where aerobic fitness level and exercise intensity seem to play a role in determining VO₂ kinetics changes in response to exercise training.

Within the limitation of our cross-sectional experimental design, the VO₂ kinetics of maximal exercise seems to follow the same pattern of that of sub-maximal response across the three levels of aerobic fitness studied. This similar pattern between these two exercise intensities is not obvious, since VO₂ kinetics during high-intensity exercise has been described to present early improvements in response to training interventions along with no concomitant changes in moderate exercise VO₂ kinetics (Krustrup et al., 2004). Such disparity has been understood as a phenomenon mainly related to the difference in exercise intensity, since a reduced reliance on non-oxidative energy source may improve cell homeostasis and exercise tolerance during, particularly, high-intensity exercise (Jones et al., 2003). Our data suggest that above a certain level of aerobic fitness, this disparity between exercise intensities may no longer exist.

Phillips et al., 1995 have demonstrated that low fit adult subjects have their VO₂ kinetics improved by exercise training. Along with this result, middle age subjects also present similar improvements in VO₂ kinetics early after training program onset (Fukuoka et al., 2002). Interestingly, Carter et al., 2000 studying active college students (VO₂max ~55 ml·kg·min^-1) did not find any improvement in moderate exercise VO₂ kinetics after 6 weeks of exercise training, although other indexes of aerobic fitness were improved (lactate threshold and VO₂max). It seems that these findings rise two questions: Does VO₂ kinetics only improve initially while aerobic fitness is rather low? And, will longer and severe training regime further improve VO₂ kinetics?

Addressing the later question, our assessment of elite-level athletes seems to provide evidence that VO₂ kinetics stop improving above a given level of aerobic fitness during both, maximal and moderate exercise. This result is in accordance with previous time course study of VO₂ kinetics response to exercise training and the similarity of MRT reported here (Table 2) and by other studies involving highly fit subjects (Koppo et al., 2004; Kilding et al., 2007). Hence, this value of MRT may represent the fastest speed that VO₂ kinetics can be.

Some studies have reported significant relationship between VO₂max and VO₂ kinetics speed (Powers et al., 1985; Gurd et al., 2005). In spite of significant correlation coefficients between VO₂max and MRT at both exercise intensities when all the groups are joined together (Figure 1 and 2), their R² are low and the difference in VO₂max along with no associated difference in MRT between the intermediate and high aerobic fitness groups (Table 2), suggest a loose association between these variables. The significant correlations are probable due to the heterogeneity and the inclusion of low fit subjects whose VO₂ kinetics seems still sensitive to changes in aerobic fitness level.

• Currently, it is reasonable to believe that the rate-limiting step of VO₂ kinetics depends on exercise intensity.
• The well known physiological adaptations induced by endurance training are likely the most extreme means to overcome rate-limiting steps determining VO₂ kinetics across exercise intensities.
• However, exercise adaptation leading individuals to the high-end of aerobic fitness level range (VO₂max > 65 ml.kg.min-1) is not able to further improve VO₂ kinetics during both, moderate and maximal intensity exercise.