The peak oxygen uptake (̇VO_2peak) and oxygen uptake kinetics are key parameters of aerobic fitness (Whipp et al., 1981) and have been typically used to determine adaptations through endurance training. The ̇VO_2peak represents the maximum rate to resynthesize ATP aerobically, whereas oxygen uptake kinetics provide an additional non-invasive insight into the integrated capacity of oxygen utilization and transport to the contracting myocytes at the onset of exercise (Poole and Jones, 2012). Furthermore, a close coupling between muscle PCr breakdown and oxygen uptake kinetics was observed at the on- and offset of exercise in children (Barker et al., 2008) and adults (Rossiter et al., 2002). Recent studies have shown that there are no differences in ̇VO_2peak between untrained children and adults when a ratio standard (mL·min^-1.kg^-1) was used (Fawkner et al., 2002; Leclair et al., 2013). However, faster oxygen uptake kinetics have been reported in children and therefore it was suggested to consider both ̇VO_2peak and oxygen uptake kinetics to determine aerobic fitness (McNarry and Jones, 2014; McNarry and Barker, 2018). Recent studies showed a faster time constant (𝜏) during moderate-intensity exercise in untrained children compared with adults (Armstrong and Barker, 2009; Fawkner and Armstrong, 2003; Fawkner et al., 2002; Leclair et al., 2013). Moreover, a faster 𝜏 with no or a significantly lower ̇VO₂ slow component (SC) was found during high-intensity exercise in children compared to adults (Armon et al., 1991; Williams et al., 2001). In addition, longitudinal studies reported an increased 𝜏 and SC over a 2-year period in children and adolescents at heavy-intensity exercise (Breese et al., 2010; Fawkner and Armstrong, 2004). It is evident that training affects the parameters of oxygen uptake kinetics. In adults, cross-sectional studies showed a faster 𝜏 in trained compared to untrained subjects irrespective of exercise intensity and type of exercise (Caputo et al., 2003; Koppo et al., 2004). In addition, a six week (Jones and Carter, 2000) and eight week (Billat, 2000) endurance training in untrained adults, reduced the amplitude of the SC at the same absolute treadmill speed. Furthermore, within highly trained athletes, those with a higher ̇VO_2peak showed a faster 𝜏 and a reduced SC (Ingham et al., 2007; Powers et al., 1985). Therefore, it can be assumed that a high training status as a result of a higher accumulated training volume (i.e. years of training), is associated with a faster 𝜏 and reduced amplitude of the SC (Poole and Jones, 2012). Whilst the effect of endurance training volume on parameters of oxygen uptake kinetics is well investigated in adults, relatively less is known about the responses in children and youth. Whilst early studies indicated no effect of training on oxygen uptake kinetics in prepubertal children (Cleuziou et al., 2002; Obert et al., 2000), more recent studies reported a significant effect of training. It has been shown, that trained children and adolescents showed faster adjustments in the oxidative response at the onset of exercise than their untrained counterparts irrespective of sex (Marwood et al., 2010; McNarry and Jones, 2014; McNarry et al., 2011a; McNarry et al., 2011b; Unnithan et al., 2015; Winlove et al., 2010). According to the results examining youth and adults, it appears that training status (i.e. endurance training history) is a potential factor to positively affect oxygen uptake kinetics and therefore exercise tolerance in both, youth and adults.

Whilst the majority of studies compared oxygen uptake kinetics between untrained children and adults, comparisons of the ̇VO₂ responses between endurance trained youth and adults have not been examined. Additionally, an influence of work rate on the ̇VO₂ response is reported in adults (Poole and Jones, 2012). Although some results indicate, that the ̇VO₂ response is independent of work rate in children no studies specifically investigated the influence of work rate on the ̇VO₂ response in children (McNarry, 2019).

The main purpose of the study was to compare oxygen uptake kinetics between endurance trained youth and adult cyclists. The second purpose was to determine differences in the oxygen uptake response at different exercise intensities within groups. We hypothesized, that the primary response would be faster in adult cyclists due to the higher training status and would be affected by exercise intensity within both groups. During heavy-intensity exercise, we expected a larger SC in adult than in youth cyclists.

Nine adult male and four female (n = 13; mean ± SD; age: 23.2 ± 4.8 years; stature: 1.76 ± 0.09 m; body mass: 68.1 ± 10.0 kg) and ten youth male and three female (n = 13; mean ± SD; age: 14.3 ± 1.5 years; stature: 1.65 ± 0.12 m; body mass: 54.2 ± 11.7 kg) members of the Austrian Cycling Federation volunteered to participate in this study. The youth cyclists had a training history of 3-4 years and trained for 10 to 12 h per week at the time of the study. Adults training history was at minimum 10 years with 15 to 20 h of training per week. Before entering the study, all participants and the guardians of the youth cyclists were informed of the experimental procedures. The youth cyclists provided assent to participate and their guardians, as well as the adult cyclists, gave written informed consent. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the institutional review board. Participants were instructed to arrive at the laboratory in a rested and fully hydrated state and to refrain from strenuous exercise in the 24 h prior each test.

Participants visited the laboratory on two occasions separated by 3-7 days. All participants performed a ramp-exercise test on a Cyclus-2 cycle ergometer (RBM Electronics, Leipzig, Germany) during the first visit. At the second visit participants performed two square-wave exercise transitions each at moderate and heavy-intensity exercise to obtain a 95% confidence interval in the primary response 𝜏 of ± 5 s. During all trials pulmonary ̇VO₂ was measured continuously and laboratory temperature was held constant at 22° with a relative humidity of ~42%.

Participants performed the ramp-exercise test at a cadence of 90-100 rev·min^-1 until exhaustion to determine maximal measures of power output (P_max) and ̇VO_2peak. Participants used their own bicycle which was mounted on a Cyclus-2 ergometer. After a 3-minute standardized warm-up at 60 W the work rate was increased continuously by 20 W·min^-1 until exhaustion. Pulmonary ventilation and gas exchange were measured continuously throughout the test via breath-by-breath open circuit spirometry (MetaMax 3B, Cortex Biophysik, Leipzig, Germany). The gas analyzer was calibrated before each test with gases of known concentrations (4.99 Vol% CO₂, 15.99 Vol% O₂, Cortex Biophysik, Leipzig, Germany). Volume and flow were calibrated with a 3-L syringe (Type M 9474-C; Cortex Biophysik, Leipzig, Germany). The participants wore a facemask and breathed through a low-resistance impeller turbine. Achievement of ̇VO_2peak was assumed as the highest 10-s average value attained before volitional exhaustion. In addition, the ventilatory threshold (VT) was determined as sub-maximal performance correlate. For determination of VT the criteria of an increase of the ventilatory equivalent of O₂ (̇VE/̇VO₂), without a concomitant increase of the ventilatory equivalent of CO₂ (̇VE/̇VCO₂) and the first loss of linearity in pulmonary ventilation (̇VE) and carbon dioxide ventilation (̇VCO₂) were used (Beaver et al., 1986).

The trials were conducted at a cadence of 90 rev·min^-1 for all intensities. The work rate for moderate-intensity exercise was set at 90% of VT and the work rate for heavy-intensity exercise was set at the difference of 50% between VT and P_max (∆50%). After a 5-min warm-up period (i.e. cycling at 40 W), each trial started with 3-min baseline cycling at a power output of 60 W followed by 6 min cycling at moderate intensity separated by 10 min cycling at 60 W before 6 min cycling at heavy intensity. The protocol was repeated after a passive rest of 1 h (Nimmerichter et al., 2020). During all trials gas exchange was measured as described above.

The ̇VO₂ breath-by-breath data were examined to exclude errant breaths from sighing, swallowing, coughing, etc., and breaths lying more than four standard deviations from the local mean of 5 data points were removed. The filtered data were linearly interpolated to provide second-by-second values for each participant. In addition, the identical transitions from each participant were ensemble-averaged and subsequently time aligned to the onset of the exercise (Whipp and Rossiter, 2005). The cardio-dynamic or phase I response was visually identified and the first 15-20 s of the transition from baseline to the target work rate were excluded from further analyses (Hebestreit et al., 1998). A single-exponential equation was used to model the kinetics of the primary (phase II) response of ̇VO₂:

where ̇VO₂ (t), ̇VO_2baseline, Amp, TD and 𝜏 represent the ̇VO₂ at any given time (t), the ̇VO₂ at the last 60 s of baseline exercise, the amplitude from baseline to its asymptote, the time delay and the time constant of the response, respectively.

According to the methods of Rossiter et al. (2002) a purpose-designed software (LabView 6.1, National Instruments, Newbury, UK) was used to identify the optimal fitting window of the primary parameter response. Beginning from the initial 60 s of exercise, the fitting window was increased iteratively by 5-s to end-exercise. For ∆50% trials, each fitting window was plotted against time of the estimated 𝜏 and through visual inspection the onset of the SC was determined as the point at which the estimated 𝜏 progressively increased following an initial plateau. The parameter estimates were then resolved by least-squares non-linear regression (Graph Pad Prism 8.1, San Diego). The amplitude of the SC was calculated as the difference between the mean of the final 30 s at end-exercise and the asymptote of the fundamental response. The ̇VO₂ gain of the primary response and of the end-exercise, was calculated by dividing the respective amplitudes by the incremental work rates above baseline (∆̇VO₂/∆WR). In addition, mean response time (MRT) was calculated as 𝜏 + TD.

Statistical analyses were performed using SPSS statistics 23 (IBM Corporation, Armonk, NY). Descriptive data are summarized as mean ± standard deviation (SD). After testing the assumption of normality using the Shapiro-Wilks’ test, a two-way repeated-measure ANOVA with Fisher’s LSD post-hoc test compared differences between groups (youth vs. adults) and intensities (moderate vs. heavy). In addition, an unpaired t-test was conducted to compare data obtained from the laboratory ramp-exercise test as well as the SC of the heavy-intensity exercise trials. The level of significance was set at p < 0.05.

The results of the ramp-exercise test and calculated target work rates for the square-wave trials are presented as absolute and normalized to body mass values in Table 1. There was a significant difference in all measures between youth and adult cyclists except the target work rate for 90% VT which was not significantly different between groups (p = 0.408) when normalized to body mass.

The parameter estimates of the transitions to moderate- and heavy-intensity exercise for both groups are provided in Table 2. ANOVA revealed significant main effects of intensity for baseline ̇VO₂ (F_{1; 24} = 8.9, p = 0.006), amplitude (F_{1; 24} = 319.5, p < 0.001), 𝜏 (F_{1; 24} = 29.9, p < 0.001), MRT (F_{1; 24} = 11.1, p = 0.003), end-exercise ̇VO₂ (F_{1; 24} = 303.2, p < 0.001) and end-exercise ̇VO₂ gain (F_{1; 24} = 4.8, p = 0.039). No significant main effect of intensity was found for TD (F_{1; 24} = 2.8, p = 0.106) and phase II ̇VO₂ gain (F_{1; 24} = 0.9, p = 0.356). Significant group differences were found for baseline ̇VO₂ (F_{1; 24} = 13.4, p = 0.001), phase II ̇VO₂ gain (F_{1; 24} = 39.7, p < 0.001), end-exercise ̇VO₂ (F_{1; 24} = 11.6, p = 0.002), end-exercise ̇VO₂ gain (F_{1; 24} = 16.1, p = 0.001) and SC (p < 0.001). No significant differences were found for amplitude (F_{1; 24} = 3.0, p = 0.058), 𝜏 (F_{1; 24} = 1.3, p = 0.263), TD (F_{1; 24} = 2.5, p = 0.124) and MRT (F_{1; 24} = 0.03, p = 0.870). In addition, no significant interactions between intensity and group were found (p = 0.294 to p = 0.717). The mean ̇VO₂ response for both intensities is illustrated in Figure 1.

The main finding of this study was that for moderate- and heavy-intensity exercise the phase II of the primary response is not significantly different between trained youth and adult cyclists. This is in contrast to our hypothesis that the primary response would be faster in adults and to studies comparing children and adults. Additionally, in both groups a SC was observed, with a significantly higher SC amplitude in adult cyclists compared to youth cyclists which is in turn comparable to results from studies comparing youth and adults.

However, in the present study adult cyclists with a longer training history and higher training volume, showed significantly higher maximal and sub-maximal physiological characteristics for both, absolute and normalized values than youth cyclists. The differences in endurance performance in the present study are not surprising. It is well documented that prolonged endurance training over several years results in increased endurance performance. Therefore, the higher endurance performance in adult cyclists can be mainly explained by the differences in training history and consequently a higher accumulated training volume.

We observed no significant differences in the phase II at the primary response of oxygen uptake kinetics at both intensities. Nevertheless, the ̇VO₂ gain in the present study was significantly different between groups, with a higher ̇VO₂ gain in youth, which is comparable to studies where untrained children showed a significantly higher ̇VO₂ gain than adults (Armon et al., 1991; Fawkner and Armstrong, 2003; Williams et al., 2001). By comparing youth and adults, the higher ̇VO₂ gain in youth is associated with a higher capacity of oxidative phosphorylation and/or a limited ability to generate ATP through anaerobic metabolism (Armstrong and Barker, 2009).

Irrespective of age the positive influence of training status on oxygen uptake kinetics has been shown in recent studies (Koppo et al., 2004; Marwood et al., 2010; Winlove et al., 2010). Furthermore, studies investigating oxygen uptake kinetics within trained groups showed that those with a higher ̇VO_2peak displayed a faster 𝜏 (Ingham et al., 2007; Powers et al., 1985). The results of the present study regarding 𝜏 are in contrast to the mentioned studies, thus an influence of age is most likely present. However, in the present study the differences in the ̇VO₂ gain for the primary response are in line with previous studies, where youth or club level athletes showed a significantly higher ̇VO₂ gain than adults or elite athletes (Ingham et al., 2007; Williams et al., 2001). In addition, we observed a SC in youth cyclists which was significantly smaller compared to adult cyclists, whereas the end-exercise ̇VO₂ gain was higher in youth cyclists. The differences in SC are most likely due to the significantly greater absolute and relative work rate performed by adult cyclists during heavy-intensity exercise.

Longitudinal and cross-sectional studies showed a slowing of 𝜏 and an increasing SC at heavy-intensity exercise with age (Breese et al., 2010; Fawkner and Armstrong, 2004; Williams et al., 2001). The mechanisms related to these changes with age are poorly understood, but it was suggested that children compared to adults, display a higher potential for oxidative metabolism (McNarry and Barker, 2018). Leclair et al. (2013) found a faster ̇VO₂ 𝜏 and a faster oxygen extraction (deoxy-hemoglobin) as measured by near-infrared spectroscopy in children compared to adults. It appears, that the enhanced oxidative metabolism in children includes a faster oxygen delivery and a higher oxygen extraction in the exercising muscles (Leclair et al., 2013; McNarry and Jones, 2014) and this in turn results in a reduced PCr breakdown at the onset of exercise. The results in the present study are comparable to studies evaluating groups at different training status and contrary to studies evaluating the effect of age differences. Based on the findings of recent studies we assume that the results in the present study are more related to long-term training rather than age-related adaptations.

In the present study the ̇VO₂ response is work rate dependent, with a significantly slower 𝜏 at heavy than in moderate-intensity exercise in trained youth and adults. This results are in line with recent studies, where untrained and trained adults displayed a slowing of the 𝜏 with increasing work rate (Koppo et al., 2004). The novel finding in the present study is, that in trained youth the slowing of the 𝜏 with increasing work rate is also present. This indicates, that in trained youth the ̇VO₂ response displays a similar behavior, as it was reported in adults.

The present study reveals novel insight in oxygen uptake kinetics of trained youth and adult cyclists. However, some limitations should be addressed. An untrained control group of youth would have been beneficial to clearly identify age-related differences. Additionally, a group of adult cyclists with a comparable training status to our group of youth cyclists would have been beneficial to examine the effect of training state. As a result of the limited number of available subjects in these categories the small sample size should be considered as a limitation. Although we applied the commonly used delta concept to ensure appropriate intensity for heavy-intensity exercise, it should be noted that the differences found should be interpreted with caution due to the suggestion that the delta concept is less secure than the use of critical power to investigate this intensity domain (Armstrong and Barker, 2009).

The results of the current study are contrary to the existing knowledge about differences in the primary response of oxygen uptake kinetics between children and/or youth and adults. Our results showed that at moderate and heavy-intensity exercise trained youth and adult cyclists displayed a similar behavior in the primary response of oxygen uptake kinetics. However, irrespective of work rate and training status, recent studies and the findings of the present study showed that the ̇VO₂ gain of the primary response and end-exercise ̇VO₂ gain in youth is higher than in adults, whereas adults showed a higher SC than youth cyclists. We assume that long-term training adaptations influence oxygen uptake kinetics and can be a potent mediator to reduce the age-related slowing of oxygen uptake kinetics.