Sport Sciences for Health https://doi.org/10.1007/s11332-018-0457-y
ORIGINAL ARTICLE
V˙O2 plateau in treadmill exercise is not dependent on anaerobic capacity Cory M. Scheadler1 · Nicholas J. Hanson2 Received: 15 January 2018 / Accepted: 18 April 2018 © Springer-Verlag Italia S.r.l., part of Springer Nature 2018
Abstract Introduction A plateau in oxygen consumption (V˙O2 plateau) remains the definitive criteria for establishing that maximal oxygen consumption (V˙O2max) was achieved during maximal exercise despite its inconsistent presence. Anaerobic capacity may assist in achieving a V˙O2 plateau. Purposes To determine if a correlation exists between maximal accumulated oxygen deficit (MAOD) and magnitude of V˙O2 plateau during treadmill exercise. Methods Participants completed submaximal, maximal, and supramaximal exercise treadmill tests on separate occasions. MAOD was determined during the supramaximal test by calculating the difference between oxygen consumption (V˙O2) and predicted V˙O2 based on extrapolation of submaximal data. V˙O2 plateau was determined as a < 50 ml change in V˙O2 over the last 60 s of exercise. Results Ten of seventeen participants showed a plateau in V˙O2. Oxygen deficit was not different between the plateau and non-plateau groups (44.2 ± 10.8 vs. 44.8 ± 8.5 ml·kg−1, p = 0.906). Oxygen deficit was not correlated with change in V˙O2 (r = − 0.087, p = 0.739). Aerobic training hours·week−1 were higher in the plateau group than non-plateau (4.8 ± 1.5 vs. 1.7 ± 2.1, p = 0.003) and correlated with change in V˙O2 (r = − 0.418, p = 0.048). Conclusions In treadmill running, there is not a significant correlation between V˙O2 plateau and MAOD as has been seen in cycling. Keywords MAOD · Oxygen consumption · Anaerobic · Plateau · Aerobic · Running
Introduction The most important criterion for having achieved maximal oxygen consumption (V˙O 2max) during a treadmill exercise test has classically been identified as a plateau in oxygen uptake (V˙O2) and has dated back to the original experiments by Hill and Lupton [1]. However, a plateau in V˙O2 has generally been reported as occurring in low incidence [2]. Secondary criteria (e.g., heart rate, respiratory exchange ratio, blood lactate) for identifying maximal effort have come under scrutiny due to their lack of validity [3]. * Cory M. Scheadler
[email protected] 1
Human Performance Laboratory, Department of Kinesiology and Health, Northern Kentucky University, 100 Nunn Dr., Highland Heights, KY 41099, USA
Department of Human Performance and Health Education, Western Michigan University, Kalamazoo, MI 49008, USA
2
Therefore, if the plateau in V˙O2 remains the only definitive way to determine whether the V˙O2 obtained is maximal, it is important to find out why some individuals exhibit a plateau and others do not. Research investigating supramaximal treadmill running (running at work rates that require energy release beyond what is sustainable from oxidative pathways alone) suggests that oxygen deficit is tied to exhaustion and has a finite limit for a given individual and fitness level [4]. The oxygen deficit, if large enough, may allow individuals to exercise until V˙O2 plateaus. The dependence of V˙O2 plateau on anaerobic capacity has been suggested before [5]. A recent study has suggested that well-trained cyclists with greater anaerobic capacities are more likely to achieve a plateau in V˙O2 during cycling exercise than those with lower anaerobic capacities [6]. Gordon et al. showed that the slope of the last 60 s of V˙O2 in cyclists that had experienced a plateau (defined as ≤ 50 ml increase in V˙O2 between the last two 30 s time intervals) was significantly, negatively related
13
Vol.:(0123456789)
Sport Sciences for Health
to maximal accumulated oxygen deficit (MAOD). Recent data has suggested that anaerobic capacity exists in ample amounts at exhaustion during incremental exercise [7] which may suggest no direct relationship exists between MAOD and V˙O2 plateau. Additionally, Gordon et al. reported that treadmill plateau incidence is not consistent with cycle ergometry plateau incidence when utilizing a group of recreationally trained males [8]. It appears that plateau incidence may not be correlated with MAOD or that in treadmill ergometry, it may only exist in very well-trained individuals. It should also be noted that although plateau was analyzed using the slope, groups were formed using the difference in V˙O2 between the last two 30 s intervals. The change in V˙O2 may be a simpler means of analyzing plateau and determining correlation with MAOD. The method used to determine the existence and magnitude of plateau may play a role in the potential correlation with MAOD. Our purpose was to determine if a significant, positive correlation exists between anaerobic capacity and the incidence and magnitude of V˙O2 plateau during a maximal exercise treadmill test in participants with a range of fitness levels. We also propose to compare two methods for determining the existence and magnitude of plateau. Our hypothesis was that MAOD would be positively correlated with both incidence and magnitude of plateau during treadmill running. We further hypothesize that both methods would result in similar incidence and magnitude of plateau.
Methods Subjects Participants consisted of 17 healthy, college-aged males with varying exercise experience ranging from recreational resistance-trained individuals to experienced marathoners and triathletes. Participant demographics can be seen in Table 1. All participants provided written consent to the risks and benefits of the study before participation. The study was approved by the Institutional Review Board at the university
Table 1 Participant demographics, listed as M ± SD Variable
Plateau
Non-plateau
All
Age (years) Height (cm) Weight (kg) BMI Aerobic running (h week−1)
24.0 ± 4.9 179.2 ± 6.2 73.5 ± 6.4 22.9 ± 1.7 4.8 ± 1.5
22.6 ± 4.5 179.2 ± 9.3 73.6 ± 10.7 22.9 ± 2.7 1.7 ± 2.1*
23.4 ± 4.7 179.2 ± 7.3 73.6 ± 8.1 22.9 ± 2.1 3.5 ± 2.3
*Significantly different from plateau (p < 0.01)
13
and followed ethical guidelines set forth in the 1964 Declaration of Helsinki.
Design and protocol Participants were asked to come to the lab on three separate days. Each trial was completed on the same laboratory treadmill (Trackmaster TMX425C, Full Vision Inc., Newton, KS). For each laboratory visit, all participants were given instructions not to exercise the day of testing, to exercise lightly the day before and not to consume food or drink within 3 h of testing. Trials were separated by at least 48 h and were all completed within 2 weeks. The first visit was to familiarize participants to the equipment and to complete multiple work rates of submaximal exercise to establish V˙O2-work rate relationships. Participants exercised at a grade of 8.0% and speeds of 2.4, 4.8, 7.2, and 9.7 km h−1, each for four minutes. A work to rest ratio of 1:1 was employed. An 8.0% incline was used to match the supramaximal exercise test. Steady-state V˙O2 was confirmed visually via graphical representation of the data. Work rate was calculated as vertical power by multiplying the participant’s body mass by the rate of vertical movement (speed × grade). The average V˙O2 over the last minute of each stage was used to determine V˙O2 work rate relationships. For each participant, a regression line was plotted for work rate and V˙O2. The regression line was forced through a y-intercept of 5.1 ml kg−1 min−1 [9]. On the second visit, participants completed a maximal incremental exercise treadmill test to volitional exhaustion to measure V˙O2peak. Treadmill speed remained constant, while incline started at 2.0% and was increased by 2.0% every 2 min. For participants who were experienced runners, treadmill speed was faster than a typical training pace but slower than tempo pace. For recreational participants, the treadmill speed was set to 9.7 km h−1. Rating of perceived exertion (RPE) and capillary blood lactate was assessed upon exhaustion. Participants immediately sat down upon termination of exercise. Post-exercise capillary blood lactate was measured immediately after exercise and then every 2 min for 10 min. On the third visit, participants completed a supramaximal exercise test for the measurement of MAOD. Participants began with a 5-min warm-up at a self-selected speed and then moved directly into another 5-min running period at a work rate estimated to elicit 85% of V˙O2peak. Participants were given 10 min to recover before beginning the supramaximal test. Each supramaximal test was completed at 8.0% incline with participants stepping onto a moving belt to signify a square wave bout of exercise. The work rate was chosen individually for each participant, estimated to fully exhaust the participant in two to four minutes [10] based on past exercise experience and results from the maximal
Sport Sciences for Health
incremental exercise test. The resultant work rates were, on average, the equivalent of 111.1% of V˙O2peak. All participants were provided with similar verbal encouragement throughout testing by the same researcher.
Methods Expired gases were analyzed using a True One 2400 metabolic cart (ParvoMedics, Sandy, UT). Flow calibration was conducted according to manufacturer’s instructions. Gas analyzers were two-point calibrated with a reference gas of known quantity (e.g., 16% O2 and 4% C O2) and room air (assumed 20.94% O 2 and 0.03% C O2) according to manufacturer’s recommendations.
Data analysis Gas analysis was completed using 30-s retrograde averages as previously described [11]. Thirty seconds was chosen for sampling interval based on recommendations in the literature [12]. A plateau in V˙O2 was accepted as a change in V˙O2 between the last two 30-s periods (ΔV˙O2) of < 50 ml min−1 [6], and this difference was used as the first plateau method. Additionally, a 9-s moving average was applied to the breath-by-breath data from the last 60 s of oxygen uptake and a line of best fit applied to determine the magnitude of the plateau (60-s slope) [6] was used at the second plateau method. RPE was assessed using the Borg 15-point scale. RER was defined as the highest 30 s value calculated by the metabolic system. Blood lactate was measured at the fingertip with a YSI 1500 Sport lactate analyzer (Yellow Springs Instruments, Yellow Springs, Ohio). On each day of testing, the analyzer was calibrated according to manufacturer instructions. Secondary criteria for maximal effort were considered as the following: RPE of ≥ 19, RER ≥ 1.15 and blood lactate ≥ 8 mmol l−1. Anaerobic capacity was estimated using the measurement of MAOD [13]. The regression line for each participant’s V˙O2-work rate relationship was extrapolated to the work rates of the supramaximal exercise test to determine its estimated V˙O2 requirements. The estimated V˙O2 requirements and attained V˙O2 for each 15-s period were time corrected by multiplying by 0.25, or other appropriate decimal when the final period was not 15 s, to get units of ml kg−1. The difference between the estimated V˙O2 requirement and the attained V˙O2 for each period of the supramaximal exercise test was summed and considered to be the MAOD.
Statistics Differences in demographics, secondary criteria for maximal effort, maximal and supramaximal exercise test
physiological variables of interest, and plateau criterion between participants who presented with a V˙O2 plateau (PL) and those who did not (NPL) were determined using independent samples t tests when data was normally distributed with equal variances. Additionally, each 60-s slope was assessed in its ability to determine plateau by determining whether it differed from zero using a one-sample t test. Normality of data was determined visually with histograms. Levene or Brown–Forsythe procedures were used to determine difference in variance based on normality of data. When unequal variances were present, a Welch’s t test was used to determine group differences. When data was not normally distributed, ranked scores were calculated and analyzed for group differences. Pearson correlation was used to identify a correlation between anaerobic capacity and the magnitude of the plateau by analyzing MAOD against the 60-s slope and ΔV˙O2. Significance was set a priori at p < 0.05. Statistical analyses were completed using SPSS version 24 (SPSS, Chicago, Illinois).
Results The duration of the maximal exercise test was 603 ± 105 s. Ten out of seventeen participants achieved a V˙O2 plateau according to the defined criteria of a ΔV˙O 2 of < 50 ml min−1. However, all participants achieved a maximal effort, defined as achieving at least one of the secondary criteria thresholds (Table 2). V˙O2peak ranged from 40.8 to 78.0 ml kg−1 min−1, confirming recreational to welltrained aerobic fitness levels (Table 3). The duration of the supramaximal exercise test was 159 ± 37 s. The five plotted points (including forced y-intercept of 5.1 ml kg−1 min−1) in the V˙O2-work rate relationship resulted in an average R2 for the regression line of 0.9571 ± 0.0196. There was a significant difference (p < 0.05) in ΔV˙O2 between PL (− 28.5 ± 40.0 ml·min−1) and NPL (176.1 ± 107.5 ml·min−1); this is expected as this was the criteria for which the groups were categorized. There was no difference (p = 0.074) in 60-s slopes between Table 2 Peak secondary criteria at maximal effort, listed as M ± SD and range Variable
Plateau
Non-plateau
RPE
19.1 ± 1.5 15.0–20.0 1.12 ± 0.04 1.07–1.18 8.9 ± 1.7 5.3–11.3
19.0 ± 1.4 17.0–20.0 1.17 ± 0.04* 1.11–1.21 10.9 ± 2.2* 8.2–14.4
RER Lactate (mmol l−1)
*Significantly different from plateau at p < 0.05
13
Sport Sciences for Health
Table 3 Physiological variables at volitional exhaustion, listed as M ± SD Variable
Plateau
Non-plateau
V˙O2peak (ml kg−1 min−1) V˙Epeak (l min−1) V˙CO2peak (l min−1) HRpeak (beats min−1)
64.1 ± 8.7 154.2 ± 20.8 5.34 ± 0.82 193 ± 9
59.1 ± 10.5 151.6 ± 21.5 4.92 ± 0.98 191 ± 16
V˙O2peak peak oxygen consumption, V˙Epeak peak minute ventilation, V˙CO2peak peak carbon dioxide production, HRpeak peak heart rate
400
correlation between the aerobic running hours and ΔV˙O2 for the study sample using a one-tailed Pearson correlation; there was a significant negative correlation (r = − 0.418, p = 0.048). Differences in peak lactate and RER also existed between groups (Table 2). The average MAOD value was 44.4 ± 9.7 ml kg −1, similar to values seen in prior research [4]. There was no difference in MAOD between PL and NPL (44.2 ± 10.8 vs. 44.8 ± 8.5 ml kg−1, p = 0.906). There was no correlation between ΔV˙O2 (55.7 ± 126.5 ml min−1) and MAOD (r = − 0.087, p = 0.739).
Discussion
-1
∆VO 2 (ml∙min )
300 200 100 0 -100 -200 -300
-200
-100
0
100
200
300
400
-1
60-sec Slope (ml∙min )
Fig. 1 ∆V˙O2 versus 60-s slope. Open squares represent PL participants. Open circles represent NPL participants. Arrows highlight participants classified differently based on plateau criteria
PL (− 18.2 ± 100.5 ml·min −1) and NPL (109.2 ± 173.7 ml·min−1). Additionally, the 60-s slopes of PL and NPL were not different than zero (p = 0.581; p = 0.147, respectively). Figure 1 shows that two NPL participants had 60-s slopes less than zero (open circles with arrows) and three PL participants (open squares with arrows) had 60-s slopes greater than zero, suggesting inconsistent determination of plateau between the two methods. If the two methods were consistent, the PL should have presented with a 60-s slope not different from zero, whereas NPL would have a 60-s slope different than zero. Due to the lack of difference of the NPL slope from zero, the potential for individual misclassifications of plateau that would occur with use of the 60-s slope (Fig. 1), and other scientific evidence suggesting a slope in V˙O2 similar to zero is not a reliable marker for maximal effort [14], we chose to remove the 60-s slope from further analysis and to solely use the ΔV˙O2 criteria for defining and measuring the magnitude of plateau. Hours of weekly running was the only significantly different demographic between PL and NPL (Table 1). Due to this finding, we sought to determine whether there was a
13
The major focus of this study was to determine if anaerobic capacity, as estimated by MAOD, was correlated with the presence of a plateau in oxygen consumption during treadmill running as has been shown in cycling [6]. A secondary focus of the study was to evaluate the relationship of MAOD with two different plateau methods. The test durations, achievement of secondary criteria by participants and strong fit of the V˙O2 work rate regression lines for individual data support proper testing procedures and data collection. The first finding in this study that needs to be discussed was the inconsistent determination of a plateau in V˙O2 between methodologies: < 50 ml increase in V˙O2 between the last two 30-s intervals (∆V˙O2) versus a slope of the last 60-s of V˙O2 data following application of data smoothing that is not different from zero (60-s slope). We recognize that the 60-s slope as an objective criterion for V˙O2 plateau has been previously suggested [15] and implemented [6]. We initially implemented the method to maintain consistency with previous research, however, to our knowledge the 60-s slope method has not been formally validated. Astorino et al. initially suggested the slope of V˙O2 versus time (e.g., 60-s slope) as an objective measure of attainment of plateau [15]. Gordon et al. initially used ∆V˙O2 to identify a plateau, while only using the 60-s slope for measurement of the magnitude of the plateau [6]. Additionally, Gordon et al. altered the method to include prior smoothing of the data using 9 s moving averages [6]. The discrepancies in use of this method highlights the fact that best practices for this method (i.e., using breath-by-breath or other sampling frequency) are lacking much like those for traditional V˙O2 sampling. We followed similar steps as Gordon et al. initially, however, we chose to measure the 60-s slopes for PL and NPL and compared them directly. We found that the use of 60-s slope as a measure of V˙O2 plateau was not justified. PL and NPL 60-s slopes were not significantly different. NPL had a 60-s slope similar to zero. Furthermore, there was a lack of consistency between methods when identifying a plateau designation for five of the participants (Fig. 1). Due to these
Sport Sciences for Health
findings, 60-s slopes were removed from analysis and only ∆V˙O2 was used. The primary finding of this study was that there was no immediate correlation between magnitude of plateau, shown as ΔV˙O2, and anaerobic capacity. MAOD did not differ between groups and did not change the ∆V˙O2 achieved at test termination. In addition to the highlighted difference in plateau method used (∆V˙O2 vs. 60-s slope), the modality difference in this study may also explain this difference in results from Gordon et al. Previous research suggested that higher anaerobic capacity increased the magnitude of plateau using trained cyclist with a cycling modality. In those not trained as cyclists, local muscular fatigue generally precludes the obtainment of a true V˙O2max [16]. In trained cyclists, exercise can continue long enough to obtain V˙O2max, likely due to the higher anaerobic capacity specific to cycling [6]. It may be that the mechanism of fatigue in treadmill exercise is not as highly related to anaerobic capacity as is typically seen in cycling. More so than treadmill exercise, exercise termination in maximal cycling exercise has traditionally been linked to local muscular fatigue. This is supported by research showing slower V˙O2 kinetics in cycling versus running, a greater anaerobic contribution in cycling versus running, and a strong relationship between exercise duration and V˙O2 obtained at exhaustion [17]. This may have allowed for a correlation to exist between MAOD and V˙O2 plateau in previous work with cycling [6]. It does not appear that a strong association between anaerobic capacity and V˙O2 plateau exists in treadmill exercise. A notable difference between PL and NPL groups is the number of hours of weekly aerobic running they participate in. The PL group was completing three times as many hours of weekly aerobic running. The running in question was not specific to workouts, particularly speed work, but focused on overall time spent running. We can only speculate how this finding may play a role in why the PL group showed a plateau in V˙O2 whereas NPL did not. Future research should focus on adaptations secondary to aerobic training, the role of aerobic training to alter V˙O2 kinetics [18] and how this may play into incidence of V˙O2 plateau. PL and NPL groups also differed in peak RER and peak lactate values, but not V˙O2, achieved at test termination. It is important to note that NPL participants would have fallen short of their V˙O2peak at the RER and lactate levels of PL participants. The difference in RER and lactate at exhaustion further substantiates previous suggestions that secondary criteria be abandoned as soon as a more valid measurement for determining maximal effort is determined. Research has shown that a high level of aerobic power does not necessarily allow for consistent attainment of a V˙O2 plateau [19], a finding that is supported by the present study in the lack of difference in V˙O2peak between PL and NPL. Based on the significantly greater amount of aerobic
training time per week for PL and the significant correlation between aerobic training volume and ΔV˙O2, we hypothesize that current aerobic training volume increases the chance for V˙O2 plateau.
Conclusion Our results indicate there is no significant direct correlation between MAOD and V˙O2 plateau during treadmill exercise for runners of varying aerobic fitness. Participants with more extensive aerobic training may be better equipped to exercise to the point of a V˙O2 plateau during treadmill exercise, verifying a true V˙O2max. Although it cannot be ruled out that large anaerobic capacities may extend exercise duration to the point of observing a plateau in V˙O2, in some individuals there are likely other mechanisms leading to fatigue prior to a plateau. Current aerobic training volume is correlated to the presence of V˙O2 plateau through yet undetermined mechanisms but may be linked to V˙O2 kinetics. It may be the case that the anaerobic capacity necessary to achieve a V˙O2 plateau is different for each individual, resulting in some individuals with high anaerobic capacities still not achieving a V˙O2 plateau. Technicians should be aware of the inconsistency in the presence of the V˙O2 plateau during maximal exercise tests and use several criteria and professional judgement when determining if exercise effort is maximal.
Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest. Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent Informed consent was obtained from all individual participants included in the study.
References 1. Hill AV, Lupton H (1922) The oxygen consumption during running. J Physiol 56:xxxii-xxxiii 2. Beltrami FG, Wong del P, Noakes TD (2014) High prevalence of false-positive plateau phenomena during V˙O2max testing in adolescents. J Sci Med Sport 17(5):526–530 3. Poole DC, Wilkerson DP, Jones AM (2008) Validity of criteria for establishing maximal O 2 uptake during ramp exercise tests. Eur J Appl Physiol 102(4):403–410 4. Carter H, Pringle JS, Barstow TJ et al (2006) Oxygen uptake kinetics during supra V˙O2max treadmill running in humans. Int J Sports Med 27(2):149–157
13
5. Wagner PD (2000) New ideas on limitations to V˙O2max. Exerc Sport Sci Rev 28(1):10–14 6. Gordon D, Hopkins S, King C et al (2011) Incidence of the plateau at V˙O2max is dependent on the anaerobic capacity. Int J Sports Med 32(1):1–6 7. Morales-Alamo D, Losa-Reyna J, Torres-Peralta R et al (2015) What limits performance during whole-body incremental exercise to exhaustion in humans? J Physiol 593(20):4631–4648 8. Gordon D, Mehter M, Gernigon M et al (2012) The effects of exercise modality on the incidence of plateau at V˙O2max. Clin Physiol Funct Imaging 32(5):394–399 9. Russell A, Le Rossignol P, Kai Lo S (2000) The precision of estimating the total energy demand: implications for the determination of the accumulated oxygen deficit. J Exerc Physiol 3(2):55–63 10. Hawkins MN, Raven PB, Snell PG et al (2007) Maximal oxygen uptake as a parametric measure of cardiorespiratory capacity. Med Sci Sports Exerc 39(1):103–107 11. Scheadler CM, Devor ST (2015) V˙O2max measured with a selfselected work rate protocol on an automated treadmill. Med Sci Sports Exerc 47(10):2158–2165 12. Robergs RA, Dwyer D, Astorino T (2010) Recommendations for improved data processing from expired gas analysis indirect calorimetry. Sports Med 40(2):95–111
13
Sport Sciences for Health 13. Medbo JI, Mohn AC, Tabata I et al (1988) Anaerobic capacity determined by maximal accumulated O2 deficit. J Appl Physiol 64(1):50–60 14. Myers J, Walsh D, Sullivan M et al (1990) Effect of sampling on variability and plateau in oxygen uptake. J Appl Physiol 68(1):404–410 15. Astorino TA, Willey J, Kinnahan J et al (2005) Elucidating determinants of the plateau in oxygen consumption at V˙O2max. Br J Sports Med 39(9):655–660 (Discussion 660) 16. Millet G (2009) Physiological differences between cycling and running. Rev Med Suisse 5(212):1564–1567 17. Sousa A, Figueiredo P, Zamparo P et al (2015) Exercise modality effect on bioenergetical performance at V˙O2max intensity. Med Sci Sports Exerc 47(8):1705–1713 18. Zoladz JA, Grassi B, Majerczak J et al (2014) Mechanisms responsible for the acceleration of pulmonary V˙O2 on-kinetics in humans after prolonged endurance training. Am J Physiol Regul Integr Comp Physiol 307(9):R1101-1114 19. Doherty M, Nobbs L, Noakes TD (2003) Low frequency of the “plateau phenomenon” during maximal exercise in elite British athletes. Eur J Appl Physiol 89(6):619–623