Society of Exercise Physiologists (ASEP) ISSN 1097-9751
An International Electronic Journal
Volume 3 Number 4 October 2000
Incidence Of The Oxygen Plateau at VO2max During Exercise Testing To Volitional Fatigue
TODD A. ASTORINO, ROBERT A. ROBERGS, FARZANEH GHIASVAND, DEREK MARKS, AND STEVE BURNS
Center for Exercise and Applied Human Physiology, The University of New Mexico, Albuquerque NM
TODD A. ASTORINO, ROBERT A. ROBERGS, FARZANEH GHIASVAND, DEREK MARKS, AND STEVE BURNS. Incidence Of The Oxygen Plateau at VO2max During Exercise Testing To Volitional Fatigue.JEPonline, 3(4):1-12, 2000. The purpose of this study was to better clarify the VO2 response to exercise to VO2maxby comparing data derived from different time averaging intervals and exercise protocols. Sixteen active subjects (12 men and 4 women, mean age, height, weight, and VO2max= 31.68.9 yr, 172.06.6 cm, 70.812.7 kg, 3,212652 mL/min) completed three different VO2maxtests on a cycle ergometer (a 25 Watt/min ramp protocol (R), a 75 Watt/3 min step protocol (S), and a 25 Watt/min ramp protocol (H) under hypoxic conditions (FIO2 = 15 %, PB = 635 mm Hg) on separate days. During each test, subjects breathed humidified air from a Tissot tank, and breath-by-breath gas exchange was obtained by a Medical Graphics metabolic cart. All breath-by-breath data were smoothed using an 11-breath moving average. These data were then time-averaged into 15, 30, and 60 s sampling intervals. Criteria for attainment of VO2maxincluded two of the following: RER > 1.1, maximal heart rate (HR) within 10 b/min of the calculated value, or an O2 plateau (VO2< 50 mL/min)with an increase in power output. Average VO2maxwas significantly lower (F (2, 30) = 84.37, p < .001) for the H (2,532562 mL/min) compared to the S (3,112660 mL/min) and R (3,212651 mL/min) protocols. Average maximal RER was significantly different (F (2, 30) = 3.77, p < .05) across protocols, however no differences were exhibited between means. Average HR at VO2maxwas significantly lower (F (2, 30) = 12.26, p < .001) during the H trial (16913 b/min) compared to the R (1769 b/min) and S (1789 b/min) protocols. The incidence at which subjects demonstrated a plateau in VO2 for all protocols combined were 100, 100, 57 and 8 % for the 11 breath, 15 s, 30 s and 1 min averaging, respectively. Data of the change in VO2 between VO2max and the closest neighboring data point revealed that variability was greatest for the longer time averaged data. This response was similar for each protocol. These findings show that shorter sampling intervals (breath-by-breath and 15 s) are most suitable for the detection of the VO2 plateau during progressive exercise to VO2max. In addition, ramp and step protocols produce similar results, and acute normobaric hypoxia does not decrease the incidence of a VO2 plateau at VO2max using 11 breath or 15 s time averaging procedures.
Maximal oxygen uptake (VO2max) is the fundamental measure of exercise physiology, and in combination with the concept of a VO2 plateau, has been described as “the single mostinfluential concept in modern exercise physiology” (1). VO2max is widely recognized as both a representation of the functional limitations of the cardiovascular system as well as a measure of aerobic fitness. This measure originates from pioneering research by A. V. Hill and colleagues (2) that not only introduced the concept but also attempted to explain its physiological mechanisms. Hill commented… “In running…the oxygen intake may attain its maximum and remain constant merely because it cannot go any higher owing to the limitations of thecirculatory and respiratory system.” Hill also attempted to explain the onset of fatigue: … “Atthe higher speeds the requirement of the body for oxygen…cannot be satisfied…lactic acid accumulates, a continuously increasing oxygen debt being incurred, fatigue and exhaustion setting in.” These conclusions were accepted by exercise physiologists for the next 70 years, and until recently the concept of VO2maxwas not critically evaluated by research.
Also introduced by Hill and colleagues in the 1920s was the concept of the oxygen plateau, or a maintenance in oxygen uptake with an increasing oxygen demand. This theory was supported by research from Taylor et al. (3) that revealed a maintenance of oxygen consumption (VO2<150 mL/min) with an increase in treadmill grade in 108 of 115 subjects who completed a discontinuous VO2maxprotocol. Nevertheless, findings by Myers et al. (4) demonstrated that only two of six subjects revealed a plateau in VO2 at termination of ramp treadmill testing. These findings refuting the O2 plateau phenomenon were corroborated by data in a follow-up study by these investigators (5), who then recommended the use of a longer sampling interval (30 - 60 s) to enhance the precision of gas exchange data.
Recently, heated debate concerning the concept of VO2maxand the O2 plateau has led to a re-examination of the assumptions and findings of classic research by Hill and Lupton. Initially, Howley, Bassett, and Walsh (6) reviewed the history and validity of current criteria for the attainment of VO2max. They concluded that none of the established criteria was suitable for confirming the appearance of VO2max. More recently, these researchers (7) responded to assertions by Noakes (8) stating that the “cardiovascular/anaerobic model” of VO2maxproposed by Hill was incorrect, and that the O2 plateau demonstrated by Hill was obtained from flawed methodology. After review of all evidence, Howley and Bassett suggested that the classical VO2maxparadigm of Hill was valid, and that Noakes’ beliefs lacked strong scientific evidence and generated several paradoxes. Not to be outdone, Noakes (1) introduced an alternative model describing additional possible limitations to maximal exercise, namely that skeletal muscle contraction is mediated by a hierarchy of regulators specifically designed to prevent organ damage consequent with myocardial ischemia.
This debate as of yet has not identified the exact limitations to VO2max, but it has challenged current exercise physiologists to reinvestigate the fundamental measure of our field and its related controls. Therefore, the primary aim of the present study was to use more precise sampling techniques (breath-by-breath sampling of VO2) to better clarify the VO2 response to exercise to VO2maxusing a variety of sampling intervals and VO2maxprotocols. It was hypothesized that shorter sampling intervals would yield greater incidence of a plateau in VO2 at VO2max in healthy, active individuals, and that this incidence would be similar for different protocols and during acute normobaric hypoxia.
Healthy, active subjects (12 males and 4 females) were recruited from the faculty, undergraduate, and graduate student populations at the university, as well as the surrounding community. The mean age, height, weight, and VO2maxof the subjects were 31.69 yr, 172.0 6.6 cm, 70.8512.67 kg, and 3,212652 mL/min, respectively. All subjects signed a written informed consent before volunteering for the study, and all procedures were approved by the university Human Subjects Institutional Review Board. All subjects resided and were acclimatized to altitudes between 1,525 and 2,225 m above sea level and all testing was completed at 1,525 m above sea level (PB = 626 - 638 mm Hg).
Each subject completed three different VO2maxprotocols, a 75 Watt/3 min step protocol (S), a 25 Watt/min ramp protocol with workload set at 50 Watt for the first two minutes (R), and the identical ramp protocol under hypoxic (FIO2 =15 % O2, PB = 635 mm Hg) conditions (H). The order of these tests was randomized according to a Latin Squares design (9) and assigned to subjects based on subject number. Subjects and all researchers with the exception of those operating the gas exchange system were blinded as to the specific VO2maxtest to be administered that day. All tests were separated by at least two days.
Initially, subjects were familiarized with the experimental apparatus, which consisted of exercise on a cycle ergometer (Corval Lode bv, Goningen, The Netherlands) during which gas exchange was continuously monitored by a breath-by-breath system (MedGraphics CPX-D, St. Paul, MN). Humidified inspired air was directed from the high-pressure tank through a sample of distilled water, and then collected in a 100 L Tissot spirometric tank. Subjects inspired air from the tank and expired through a plastic mouthpiece and 3-way valve (Hans Rudolph Inc., Kansas City, MO). This metabolic cart measures expired airflow by means of a pneumotach connected to the mouthpiece. A sample line is connected to the pneumotach from which air is continuously pumped to O2 and CO2 gas analyzers. Prior to testing, the pneumotach was calibrated with ten samples from a 3 L calibration syringe. The gas analyzers were also calibrated before each test to room air and medically-certified calibration gases (12.29 % O2 and 5.12% CO2, respectively). Heart rate (HR) was continuously recorded during exercise by electrocardiography (Quinton Instruments, Seattle, WA).
Prior to exercise, the subject’s height and weight were recorded. Seat and bar height of the cycle ergometer were set according to the subject’s specifications, and resting gas exchange data were obtained. Exercise was then started, with subjects completing the predetermined VO2maxprotocol of between 8 and 12 minutes duration. Subjects were instructed to maintain a pedal cadence between 80 and 100 rpm during exercise and to exercise to volitional fatigue. Termination of the test occurred when the subject was unable to maintain a pedaling cadence of 40 rpm.
Maximal oxygen consumption was assessed by the attainment of the following criteria: (1) a plateau (VO2< 50 mL/minat VO2peak and the closest neighboring data point) in VO2 with increases in external work, (2) maximal respiratory exchange ratio (RER) > 1.1, and (3) maximal HR within 10 b/min of the age-predicted maximum (220 – age). All subjects met the first two criteria.
Breath-by-breath gas exchange data from all tests were transferred to a spreadsheet program (Microsoft Excel 8.0) for further analysis. All VO2 data were smoothed using an 11-breath moving average. In addition, data from the VO2maxtests were time-averaged using 15, 30, and 60 s intervals to examine the incidence of an oxygen plateau at different sampling intervals.
All data were analyzed using SPSS (Version 8.0). One-way analysis of variance (ANOVA) with repeated measures was used to examine differences in VO2max, RERmax, and HRmaxbetween the VO2maxtests. If a significant F ratio was obtained, Tukey’s HSD was used to locate differences between the means. Statistical significance was set at 0.05. With a sample size equal to sixteen and standard deviation equal to 600 mL/min, we could detect a difference in VO2 of 540 mL/min using an unpaired, one-tailed t-test with significance set at 0.05.
Gas exchange and heart rate data
VO2maxwas significantly different across protocols, F (2, 30)=84.37, p < 0.001, MSe=25,617.65. Average VO2maxfor the H (2,532562 mL/min) was significantly lower (HSD=14.52, p<0.001, and HSD=17.01, p<0.001) compared to the S (3,112660 mL/min) and R (3,212652 mL/min) protocols (Figure 1a). A significant main effect for maximal RER was exhibited between protocols, F (2, 30)=3.77, p <0.05, MSe=2.60 X 10-3. However, Tukey’s post hoc test revealed no differences between mean values of maximal RER for the R (1.30.1) compared to the H (1.40.1) or S (1.30.1) protocol (Figure 1b). Maximal HR was significantly different between the protocols, F (2, 30)=12.26, p < .05, MSe=28.03. Maximal HR during H (16913 b/min) was significantly lower (HSD=6.71, p<0.01) than that in the S (1789 b/min) protocol (Figure 1c).
Figure 1: MeanSD data from the different protocols for a) VO2max, b) RERmax and c) HRmax.
* = significantly lower in H compared to S and R protocols; # = significantly lower in H compared to S protocol.
Incidence of a plateau in VO2
All subjects for all tests demonstrated a plateau in VO2 when VO2 was sampled either breath-by-breath or every 15 s (Figures 2a and 2b). Subjects expressed a plateau 57 % of the time (27 of 48 tests) when data were sampled every 30 s (Figure 2c). For the 30 s sampling interval, the R protocol displayed a plateau in VO2 during 8 of 16 VO2max tests, while the S and H protocols revealed appearance of the plateau in 6 and 13 out of 16 tests, respectively. In contrast, only 8 % of subjects (4 of 48 tests) displayed a plateau in VO2 when data were sampled every minute (Figure 2d). Table 1 presents data of VO2 between VO2max and the closest neighboring data point for all subjects across all protocols. It is evident that the VO2 at VO2max is smaller with shorter sampling intervals, and when gas exchange data are obtained breath-by-breath or every 15 s, all subjects display a VO2 at VO2maxless than 50 mL/min, indicating a plateau in VO2. Data of the mean change in VO2 between VO2max and the closest neighboring data point based on 60 s, 30 s, 15 s, and breath-by-breath data revealed that variability was greatest for the longer time averaged data (Figure 3). This response was similar for each protocol. Means of this data for the 60 s, 30 s, 15 s, and breath-by-breath averaging were 149.370.8, 55.140.7, 22.315.5, and 11.211.5 mL/min, respectively. With the exception of the H protocol, the mean VO2 at VO2maxwas greater than 50 mL/min when longer (30 s and 60 s) sampling intervals were used.
Figure 2: Appearance of the VO2 Plateau for a) 11 breath average, b) 15 s time-averaging c) 30 s time-averaging, and d) 60 s time-averaging for a representative subject.
Table 1: VO2 at VO2maxax for all Subjects across all Protocols and Data Sampling Intervals
a = VO2 in mL/min; b = Breath-by-breath sampling interval
Figure 3: The change in VO2 between VO2max and the closest neighboring VO2 data point for each trial and data averaging method. Compiled means for all protocols at each averaging interval were all significantly (p<0.05) different from each other.
Attainment of RER and heart rate criteria
Table 2 demonstrates that all subjects expressed an RER greater than 1.15 during the R protocol. In contrast, only 38 % of subjects met the HR criteria during the R protocol (their observed HRmax at VO2max was within 10 b/min of the age-predicted theoretical value (220 – age)).
Table 2: Incidence of the Attainment of HR and RER criteria at VO2max for each subject.
188.4 + 8.9
176.2 + 9.8
1.32 + .07
a = Theoretical HRmax = (220 – age); b = Observed HRmax; c = RERmax criteria > 1.1
The primary aim of the present investigation was to better clarify the VO2 response to exercise to VO2max using more accurate sampling techniques as well as a variety of VO2maxprotocols and sampling intervals. The most significant finding was that a plateau in VO2 occurs universally in young, healthy, active subjects exercising to volitional fatigue when shorter sampling intervals (breath-by-breath and 15 s) are used. Second, the incidence of the O2 plateau is independent of the specific VO2max protocol prescribed. These conclusions may influence current exercise physiologists to realize that the O2 plateau is a real phenomenon, and that its incidence is influenced by the specific sampling interval used by gas exchange acquisition systems.
The O2 plateau concept was introduced by Hill and Lupton (2) in response to an observed maintenance of VO2 with greater speeds of running in men. Data obtained three decades later by Taylor et al. (3) supported the plateau phenomenon, as it demonstrated a plateau in VO2 in 108 out of 115 subjects completing a discontinuous VO2max protocol on a treadmill. Furthermore, results from Mitchell et al. (10) revealed a plateau in VO2 at VO2max, representing an increase in VO2<54 mL/min at the highest workload, in 72 % of subjects completing consecutive but intermittent exercise bouts of 2.5 min duration. Nevertheless, subsequent research (11, 12) in children required to complete treadmill exercise to exhaustion suggested that a plateau in VO2 (< 2.1 mL/kg/min) is elusive in this population. Similar findings were reported by Rowland (13) in nine children (mean age=11.41.2 yr) and by Armstrong et al. (14) in 35 children (mean age=9.90.4 yr). More recent data pertaining to the incidence of a plateau in VO2 at VO2max is also equivocal in adults. Stamford (15) denoted a plateau in VO2 (< 150 mL/min) between successive treadmill grades in ten men completing four different VO2max protocols on a treadmill. Furthermore, Eldridge et al. (16) observed a plateau in VO2 (<3 mL/kg/min during the final 90 s of exercise) in 77 % of normal subjects and patients with coronary artery disease completing maximal exercise. In addition, other researchers (17) reported a plateau in VO2 (change in VO2<150 mL/min, or 2.1 mL/kg/min, between successive treadmill elevations) in 50 % and 60 % of active males completing continuous and discontinuous VO2max tests on a treadmill. Similar findings regarding the prevalence of a plateau in VO2 in adults have also been demonstrated by other investigators (18, 10). However, Froelicher et al. (19) observed an O2 plateau or a decline in VO2 at volitional fatigue in 33 % of subjects completing the Taylor protocol, 17 % of subjects completing the Balke protocol, and 7 % of subjects performing the Bruce protocol. Similarly, data by Myers et al. (4) revealed a plateau in VO2, defined as the slope of the VO2-time relationship not differing significantly from zero, occurred in only two of six subjects in response to ramp treadmill testing. A subsequent investigation (5) observed marked variability in the slopes of the VO2 response to both submaximal exercise and incremental exercise to fatigue, regardless of the sampling interval used, leading these authors to conclude that the O2 plateau is not a valid physiological marker for maximal exertion.
An initial explanation for the universal incidence of the O2 plateau in the present investigation compared to previous findings is the sampling technique used. Breath-by-breath gas exchange offers great temporal resolution by calculating values for VO2 and VCO2 more frequently than previously established sampling techniques. The 11-breath moving average employed herein attenuated the majority of inherent noise in breath-by-breath data, as each breath represented the average value of the five breaths following and preceding it. Thus, a very precise representation of the VO2-O2 demand relationship was obtained throughout exercise, leading to the ability to detect a VO2 plateau at VO2max if indeed it did occur. Furthermore, the longer sampling intervals proposed by previous research (5, 6) would require subjects to sustain workloads at or above VO2max for up to 2 - 3 min to demonstrate a plateau, which is not feasible in most individuals. Also, observations in our laboratory reveal that longer sampling intervals tend to significantly underestimate VO2max. In the present study, the O2 plateau was not always realized when gas exchange data were sampled every 30 s (57 % incidence) and 60 s (8 % incidence). Using 1 min data averaging techniques, Duncan et al. (17) observed non-incidence of the plateau in 45 % of subjects. This is not surprising, since such long sampling intervals could easily miss the appearance of a plateau in VO2. Similarly, Myers et al. (5) denoted a lack of a plateau in VO2 in 66 % of subjects when gas exchange was obtained every 30 s. Therefore, it is apparent that the duration of the specific sampling interval used has a dramatic effect on prevalence of the VO2 plateau.
The rather stringent plateau criterion (VO2<50 mL/min with an increase in power output) used in the present study lends further credence to our findings. In fact, it has been suggested that cut-off values of 50 - 60 mL/min constitute a true plateau as they approach the limitations of one’s ability to measure changes in VO2max (6). The criterion used by Taylor et al. (3), a change in VO2<150 mL/min or 2.1 mL/kg/min with a 2.5 % increase in grade, has been extensively used to confirm incidence of VO2max. Astrand (20) employed a mere leveling off of oxygen uptake as a suitable criterion, yet in earlier work (18) he denoted a change in VO2 less than 80 mL/min to confirm that VO2max had been attained. Also, Issekutz et al. (21) defined the O2 plateau as less than 100 mL/min, since this VO2 was about 1/3 of the oxygen uptake required for a stage change. Lastly, Cumming and Friesen (22) used a <50 mL/min criterion, since it was roughly equal to 1/3 of the anticipated increase in VO2 for a given increase in power output. Thus, the scientific literature lacks an exact definition of a plateau in VO2 at VO2max, although the criterion established in the present study seems suitable for future application in exercise testing.
Our finding of the incidence of the O2 plateau across different VO2max tests is supported by previous research. Sheehan et al. (23) required a group of 10 - 12 year-old boys to complete a series of treadmill protocols to VO2max. Results indicated that a plateau in VO2 was common across the different protocols. Moreover, Stamford (15) observed a plateau in VO2 for each of four step and constant load VO2max tests performed by ten young men. Previous investigations failing to observe a plateau in VO2 across different protocols (19, 17) are characterized by 1 min sampling intervals which we show are unable to precisely identify a plateau in VO2. Again, it is apparent that the specific sampling intervals used greatly influence the prevalence of the O2 plateau across different VO2max protocols.
Despite the fact that the concept of VO2max has existed for over 75 years, currently no suitable criteria exist to confirm its incidence. A thorough review (6) of the efficacy of a plateau in VO2, elevated RER and blood lactate concentrations, and maximal heart rate highlighted the limitations of previous dogma, namely that many of these criteria were developed from intermittent protocols with dramatic increases in power output per day. In contrast, at the present time continuous tests with smaller work rate increments per stage are widely used. Findings from the present investigation offer two suitable criteria to confirm the incidence of VO2max: (1) a plateau in VO2< 50 mL/min with an increase in power output, and (2) RER at VO2max > 1.15. The maximal HR criterion within 10 b/min of the age-predicted value was achieved in only 38 % of our subjects, implying that HR at VO2max is highly variable among individuals. Almost 30 yr ago, Cumming and Borysyk (24) stated that the range of maximal heart rate was too variable to use as a criterion to confirm VO2max. Furthermore, the American College of Sports Medicine (25) mandates that estimated maximal heart rates should not be used to identify an endpoint during exercise testing. Therefore, this criterion should not be used to confirm the attainment of VO2max.
However, the ideal statistical assessment of the suitability of criteria to confirm VO2max is not to compare mean values. Valid criteria would be characterized by not only being related to a VO2 plateau but also to discriminate between individuals who do and do not attain a plateau. To date, no research design and appropriate statistical assessment has been conducted on this topic.
In the past, the term VO2peak has been used rather interchangeably with VO2max to reflect the highest value attained for maximal oxygen uptake. Rowell (26) commented that VO2peak represents the highest VO2 attained under specific conditions, including arm or cycle ergometry, whereas VO2max describes the highest VO2 value achieved during treadmill exercise involving at least 50 % of the muscle mass. In addition, VO2max is commonly used when criteria are not satisfied, although little agreement exists as to what they really are. Also, it has been suggested that VO2peak should represent symptom-limited VO2 during exercise, and is not related to VO2max(27). All in all, our establishment of what appear to be two suitable criteria support the notion that VO2peak should not be applied to healthy subjects exercising to volitional fatigue if these criteria are met. We recommend that the term VO2peak should be used if the VO2 plateau and RER criteria are not met.
All subjects expressed a plateau in VO2 during the severe H trial when gas exchange data were obtained breath-by-breath or every 15 s. This corroborates previous work (28) denoting a plateau in one-leg VO2 at VO2max in sedentary subjects during acute hypoxia (12 and 15 %O2). Furthermore, when gas exchange data were obtained every 30 s, 13 out of 16 subjects expressed a plateau in VO2 in response to the H trial, in comparison to prevalence of the O2 plateau in only 8 and 6 subjects during the R and S protocols, respectively. Data in rowers (29) exercising to VO2max during hypoxia (15.8 %O2) also revealed a plateau in VO2 at VO2max. The prescribed hypoxia in the present study was equivalent to 3,900 m, a rather moderate to high altitude. In response to the H trial, mean VO2max was 21 % lower compared to the R trial, which is similar to the approximate 9.2 % decrease in VO2max observed with every additional 1,000 m increase in altitude above 700 m (30). However, individual VO2max decrements in response to hypoxia are highly variable. In regards to the RER criterion, Knight et al. (31) reported a mean RER equal to 1.360.03 in male cyclists exercising to VO2max in hypoxia (12 %O2). This is identical to the maximal RER (1.360.09) recorded in the present study. Taken together, these data imply that VO2max is a valid concept during hypoxic conditions, as argued by Wagner (32).
The primary finding of the present investigation was that sampling intervals dramatically influence the incidence of the VO2 plateau at VO2max. When shorter sampling intervals (11 breath and 15 s averages) were used, there was 100 % incidence of a plateau (VO2<50 mL/min) in VO2 at VO2max with an increase in O2 demand. However, the VO2 plateau was manifested only 57 % and 8 % of the time when gas exchange was obtained every 30 s and 60 s, respectively. In addition, incidence of the VO2 plateau was similar for ramp and 3 min step increment protocols, as well as during hypoxia equivalent to 3,900 m altitude. Two suitable criteria to confirm appearance of VO2max include a plateau in VO2< 50 mL/min and RER > 1.15 at VO2max. Because of the establishment of these criteria, VO2peakshould only be used to describe symptom-limited VO2max in the absence of a VO2 plateau or the RER criteria. In addition, it is recommended that relatively short sampling intervals (< 15 s) be used to assess the VO2 response to incremental exercise to VO2max. The present findings may lead scientists and clinicians to alter the methodology used in exercise testing to obtain a better assessment of VO2max.
1. Noakes TD. Maximal oxygen uptake: “classical” versus “contemporary” viewpoints: a rebuttal. Med Sci Sports Exerc 1998;30:1381-98.
2. Hill AV, Lupton H. Muscular exercise, lactic acid, and the supply and utilization of oxygen. Q J Med 1923;16:135-71.
3. Taylor HL, Buskirk E, Henschel A. Maximal oxygen intake as an objective measure of cardiorespiratory performance. J Appl Physiol 1955;8:73-80.
4. Myers J, Walsh D, Buchanan N, Froelicher VF. Can maximal cardiopulmonary capacity be recognized by a plateau in oxygen uptake? Chest 1989;96:1312-16.
5. Myers J, Walsh D, Sullivan N, Froelicher VF. Effect of sampling on variability and plateau in oxygen uptake. J Appl Physiol 1990;68:404-10.
6. Howley, ET, Bassett, DR, Welch HG. Criteria for maximal oxygen uptake: review and commentary. Med Sci Sports Exerc 1995;27:1292-01.
7. Bassett DR & Howley, ET. Maximal oxygen uptake: “classical” versus “contemporary” viewpoints. Med Sci Sports Exerc 1997;29:591-03.
8. Noakes TD. Challenging beliefs: ex africa semper aliquid novi. Med Sci Sports Exerc 1997; 29:571-90.
9. Keppel G. Design and analysis: A researcher’s handbook. London: Prentice Hall, 1983.
10. Mitchell JH, Sproule BJ, Chapman CB. The physiological meaning of the maximal oxygen intake test. J Clin Invest 1958;37:538-47.
11. Cunningham DA, van Waterschoot BM, Paterson DH, Lefcoe M, Sangal SP. Reliability and reproducibility of maximal oxygen uptake measurement in children. Med Sci Sports 1977; 9:104-08.
12. Paterson DH, Cunningham DA, Donner A. The effect of different treadmill speeds on the variability of maximal oxygen uptake in children. Eur J ApplPhysiol 1981;47:113-22.
13. Rowland TW. Does peak VO2 reflect VO2max in children?: evidence from supramaximal testing. Med Sci Sports Exerc 1993;25:689-93.
14. Armstrong N, Welsman J, Winsley R. Is peak VO2 a maximal index of children’s aerobic fitness? Int J Sports Med 1996;17:356-9.
15. Stamford BA. Step increment versus constant load tests for determination of maximal oxygen uptake. Eur J Appl Physiol 1976;35:89-93.
16. Eldridge JE, Ramsey-Green C, Hossack KF. Effects of the limiting symptom on the achievement of maximal oxygen consumption in patients with coronary artery disease. Am JCardiol 1986;57:513-7.
17. Duncan GE, Howley ET, Johnson BN. Applicability of VO2max criteria: discontinuous versus continuous protocols. Med Sci Sports Exerc 1997;29(2):273-8.
18. Astrand I. Aerobic work capacity in men and women with special reference to age. ActaPhysiol Scand 1960;49:1-92.
19. Froelicher VF, Bramwell H, Davis G, Noguera I, Stewart A, Lancaster MC. A comparison of three maximal treadmill exercise protocols. J Appl Physiol 1974;36:720-5.
20. Astrand P-O. Experimental studies of physical working capacity in relation tosex andage. Copenhagen: Ejnar Munksgaard, 1952.
21. Issekutz B, Birkhead NC, Rodahl K. Use of respiratory quotients in assessment of aerobic work capacity. J Appl Physiol 1962;17:47-50.
22. Cumming GR, Friesen W. Bicycle ergometer measurement of maximal oxygen uptake in children. Can J Physiol Pharmacol 1967;45:937-46.
23. Sheehan JM, Rowland TW, Burke EJ. A comparison of four treadmill protocols for determination of maximal oxygen uptake in 10-12-year-old boys. Int J Sports Med 1987;8:31-4.
24. Cumming GR, Borysyk, LM. Criteria for maximal oxygen uptake in men over 40 in a population survey. Med Sci Sports 1972;4:18-20.
25. American College of Sports Medicine. Guidelines for exercise testing and prescription. Philadelphia: Lea & Febiger, 1991.
26. Rowell LB. Human cardiovascular adjustments to exercise and thermal stress. Physiol Rev 1974;54:75-103.
27. Weber KT, Janicki JS, Ward DM, McElroy PA. Measurement and interpretation of maximal oxygen uptake in patients with chronic coronary or circulatory failure. J Clin Monit 1987;3:31-7.
28. Cardus J, Marrades RM, Roca J, Barbera JA, Diaz O, Masclans JR et al. Effects of FIO2 on leg VO2 during cycle ergometry in sedentary subjects. Med Sci Sports Exerc 1998;30:697-03.
29. Peltonen JE, Rantamaki J, Niittymaki SP, Sweins K, Viitasalo JT, Rusko HK. Effects of oxygen fraction in inspired air on rowing performance. Med Sci Sports Exerc 1995;27:573-9.
30. Robergs RA & Roberts, SO. Fundamental principles of exercise physiology: For fitness, performance, and health. Boston: McGraw Hill, 2000.
31. Knight DR, Schaffartzik W, Poole DC, Hogan MC, Bebout DE, Wagner PD. Effects of hyperoxia on maximal leg O2 supply and utilization in men. J Appl Physiol 1993;75:2586-94.
32. Wagner PD. Determinants of maximal oxygen transport and utilization. Ann Rev Physiol 1996;58:21-50.
Address for Correspondence:
Robert A. Robergs, Center for Exercise and Applied Human Physiology, Johnson Center, Room B143, The University of New Mexico, Albuquerque NM 87131-1251, Phone: (505) 277-2658, Fax: (505) 277-9742, Email: firstname.lastname@example.org