Criteria for Determination of Maximal Oxygen Uptake: A Brief Critique and Recommendations for Future Research

Sports Med 2007; 37 (12): 1019-1028 0112-1642/07/0012-1019/$44.95/0 CURRENT OPINION  2007 Adis Data Information BV. All rights reserved. Criteria for Determination of Maximal Oxygen Uptake A Brief Critique and Recommendations for Future Research Adrian W. Midgley,1 Lars R. McNaughton,1 Remco Polman1 and David Marchant2 1 2 Department of Sport, Health and Exercise Science, University of Hull, Hull, UK Department of Sport and Physical Activity, Edge Hill University, Ormskirk, UK Abstract Although the concept of maximal oxygen uptake (V̇O2max) was conceived as early as 1923, the criteria used to establish whether a true V̇O2max has been attained have been heavily criticised. Consequently, an improvement in the methodology of the existing criteria, or development of new criteria, is required. In order to be valid across experimental studies, new or improved criteria need to be independent of exercise modality, test protocol and subject characteristics. One procedure that has shown potential for yielding valid V̇O2max criteria is the verification phase, which consists of a supramaximal constant speed run to exhaustion performed after the incremental phase of a V̇O2max test. A peak oxygen uptake (V̇O2peak) in the verification phase that is similar (within the tolerance of measurement error, e.g. within 2%) to the V̇O2max value attained in the incremental phase would indicate that a true V̇O2max has been elicited. Verification of the maximal heart rate would also indicate that a subject has given a maximum effort. Although the validity of the present methodology for identifying an oxygen uptake (V̇O2) plateau is questionable, a V̇O2 plateau criterion based on the individual slope of the V̇O2-work-rate relationship should improve its validity. This approach also allows determination of the ‘total V̇O2 plateau’, which is in contrast to currently used V̇O2 plateau criteria that are based on the difference in V̇O2 between only two test stages or V̇O2 data points. The ratings of perceived exertion scale has been criticised for being a one-dimensional measure of physical effort and V̇O2max criteria based on a multidimensional psychophysiological approach should increase validity. Visual analogue scales can be used to assess aspects such as muscular pain, determination and overall perceived effort. Furthermore, they are easy to complete and have demonstrated good reliability and validity in clinical and health settings. Future research should explore these and other potential approaches to developing new or improved V̇O2max criteria, so that, ultimately, a standardised set of V̇O2max criteria can be established. At present, however, the greatest challenge is identifying V̇O2max criteria that remain valid across studies. Although the concept of maximal oxygen uptake (V̇O2max) was conceived as early as 1923,[1] the validity of some of the test procedures, such as the recommended test duration[2] and appropriate Midgley et al. 1020 breath-by-breath data averaging methods[3] have not been adequately supported by experimental research. Another procedure that has not undergone rigorous scientific scrutiny is the application of criteria used to establish whether a ‘true’ V̇O2max has been attained (hereafter referred to as V̇O2max criteria). The application of valid and objective V̇O2max criteria is for quality assurance during experimental research, so that invalid V̇O2max values, due to poor subject effort[4] or inappropriate test protocols,[2] do not confound the interpretation of the findings. This may be particularly important when comparing the results of cross-sectional studies, as opposed to longitudinal studies, where a consistent effort during repeated testing for the latter, may be comparatively more important. This article provides a brief critique of currently used V̇O2max criteria and makes recommendations for future research that might identify criteria that possess greater validity than those currently used. 1. Currently Used Maximal Oxygen Uptake (V̇O2max) Criteria Howley et al.[5] reported the frequency with which V̇O2max criteria were used in studies published in Medicine and Science in Sports and Exercise between October 1993 and May 1994 (table I). We surveyed studies published between October 2005 and May 2006 in the same journal (table I), which highlights that despite heavy criticism,[5-7] the use of V̇O2max criteria has changed little over the past 12 years. There appears to be no change in the type or number of criteria used, or the consistency in the threshold values used to define each criterion. However, in the original 1993–4 survey, 76% of studies reported one or more V̇O2max criteria, compared with 44% in our 2005–6 survey. The criticisms that have been directed at the V̇O2max criteria[5-7] may be the cause of this apparent reduction in their use. Small sample bias relating to these surveys, or a change in editorial decisions regarding acceptance of studies not reporting V̇O2max criteria, are also possible explanations. In an attempt to provide a broader view of currently used V̇O2max criteria, we also surveyed studies published in four  2007 Adis Data Information BV. All rights reserved. prominent (based on impact factor) sports science and applied physiology journals between August 2005 and July 2006 (table I). This larger survey further emphasises the considerable variation in currently used V̇O2max criteria, as well as the relatively high percentage (62%) of studies that do not use or report criteria. 2. A Brief Critique of Current V̇O2max Criteria Howley et al.[5] highlighted that the oxygen uptake (V̇O2) plateau, respiratory exchange ratio, and blood lactate criteria originated from experimental studies conducted over 40 years ago.[4,8,9] Several authors have suggested that since these studies used specific exercise modalities, test protocols and subjects, directly applying any of these criteria to studies using different experimental methodology and subjects is unlikely to be valid.[5,10-12] Moreover, we question the validity of the criteria in relation to the methodology and subjects used in the original studies. A V̇O2 plateau is defined as a small or no increase in V̇O2 in response to an increase in work rate[4] and is used to demonstrate that the rate of oxygen transport and utilisation has reached its limit. However, unless an absolute plateau is used (i.e. no increase in V̇O2), a V̇O2 plateau indicates only that the rate of change in the V̇O2-work-rate relationship has slowed, not that V̇O2 has reached its maximum. The peak value achieved may instead be related to the subject’s limit of exercise tolerance or level of effort, rather than a limit of oxygen transport and utilisation. The term peak oxygen uptake (V̇O2peak) has been considered more appropriate where no V̇O2 plateau is evident.[13] However, the terms V̇O2max and V̇O2peak have been used inconsistently and we believe the term V̇O2peak should be used only in clinical settings where exercise tolerance is symptom limited in conjunction with a pathophysiological condition. This suggestion is based on the observation that when subjects perform two identical V̇O2max tests, but demonstrate a V̇O2 plateau in only one of the tests, there is no appreciable difference in V̇O2max between the two Sports Med 2007; 37 (12) Maximal Oxygen Uptake Criteria 1021 Table I. Criteria used for maximal oxygen uptake (V̇O2max) determination in: 29 studies published in Medicine and Science in Sports and Exercise (MSSE) between October 1993 and May 1994;[5] 39 studies published in MSSE between October 2005 and May 2006; and 207 studies published in four prominent (based on impact factor) sports medicine and applied physiology journals between August 2005 and July 2006 Criterion Value used MSSE 1993–4 MSSE 2005–6 Four journals 2005–6 7 22 128 Unspecified 3 7 28 Absolute plateau 1 0 2 ≤2.1 mL/kg/min 4 1 10 ≤100 mL/min 0 0 5 ≤150 mL/min 3 1 14 ≤200 mL/min 0 1 2 ≤280 mL/min 1 0 0 V̇O2 less than predicted 1 0 0 ≥10 mmol/L 0 0 3 ≥8 mmol/L 1 1 6 ≥1.20 0 0 1 ≥1.15 0 0 9 ≥1.13 1 0 0 ≥1.12 0 0 1 ≥1.10 7 13 55 ≥1.08 0 0 1 ≥1.05 2 0 3 ≥1.00 4 1 5 Plateau 1 1 4 None stated V̇O2 plateau BLa concentration RER Heart rate ±5 beats/min APMHR 3 3 3 ±10 beats/min APMHR 0 1 16 ±15 beats/min APMHR 1 0 0 ≥100% APMHR 2 1 13 ≥95% APMHR 0 1 8 ≥90% APMHR 3 1 11 ≥85% APMHR 0 1 3 Within 1 SD of APMHR 0 0 1 Close to APMHR 0 0 1 ≥19 ? 1 1 ≥18 ? 2 5 ≥17 ? 0 1 V̇O2max verification <135 mL/min ? 0 1 V̇E/V̇O2 >30–35 RPE ? 1 1 Reduced pedal rate ? 3 14 Subject exhaustion ? 1 9 Lower limb fatigue ? 0 2 Signs of intense effort ? 0 2 Dyspnoea ? 0 1 APMHR = age-predicted maximal heart rate; BLa = blood lactate; RER = respiratory exchange ratio; RPE = rating of perceived exertion; V̇E = minute ventilation; V̇O2 = oxygen uptake; ? indicates not surveyed/reported by Howley et al.[5] tests.[11,14] The V̇O2 plateau criterion is particularly limited when the plateau threshold value is not relat-  2007 Adis Data Information BV. All rights reserved. ed to the expected change in V̇O2 for each stage increment for a particular individual. The V̇O2 pla- Sports Med 2007; 37 (12) 1022 teau criterion threshold of ≤150 mL/min (or ≤2.1 mL/kg/min) first introduced by Taylor et al.,[4] was based on a mean increase in V̇O2 of 299 ± 86 mL/ min per stage increment for all subjects. The standard deviation of the mean V̇O2 response indicates that the expected change in V̇O2 for each stage increment, for some subjects, would have been less than the V̇O2 plateau criterion threshold (e.g. assuming a normal distribution, 2.5% of subjects would have an expected change in V̇O2 for each stage increment <130 mL/min [299 mL/min minus 1.96 SD]). This problem is particularly evident when directly applying the V̇O2 plateau criterion of Taylor et al.[4] to test protocols incorporating small increments, where the change in V̇O2 per stage increment will be comparatively small. Conversely, using the mean subject V̇O2 response will result in the V̇O2 plateau criterion threshold being a relatively small percentage of some subjects’ change in V̇O2 for each stage increment. These subjects would less likely satisfy the V̇O2 plateau criterion. Several V̇O2max criteria are based on surpassing threshold values for the respiratory exchange ratio and heart rate during the exercise test, or threshold values for post-exercise blood lactate concentration, and are used as evidence that the subject has given a maximum effort. However, the large between-subject variation in these variables means that many subjects will satisfy these criteria during submaximal efforts. For example, Duncan et al.[7] reported a mean maximal post-exercise blood lactate concentration of 14.3 ± 2.7 mmol/L, indicating that most subjects would probably have achieved the 8 mmol/ L criterion threshold if they stopped exercising some time before reaching their limit of exercise tolerance. A large between-subject variation also means that some subjects may not satisfy a particular criterion even when a maximum effort is given. This limitation may be more pronounced for the heart rate criterion based on the attainment of a percentage of the age-predicted maximal heart rate (HRmax), since the 95% confidence interval for agepredicted HRmax has been reported to be as large as 45 beats/min.[15] The ability to satisfy particular criteria may also be subject specific. For example,  2007 Adis Data Information BV. All rights reserved. Midgley et al. endurance-trained individuals have a lower capacity for anaerobic metabolism than sprint-trained individuals,[16] and may find it more difficult to satisfy the blood lactate and respiratory exchange ratio criteria. Although we are unaware of any studies that have directly investigated this possibility, mean maximal respiratory exchange ratios for endurance athletes that are well below common threshold values for this criteria, have been reported (e.g. Meyer et al.[17]). A statistically significant decrease in the mean maximal respiratory exchange ratio in response to endurance training, and in association with an enhancement of V̇O2max and the ventilatory threshold, has also been reported.[18] Relatively limited attention has been given to the psychophysiological components associated with the determination of V̇O2max, with research primarily focusing on Borg’s concept of perceived exertion, measured using Borg’s 15-point Ratings of Perceived Exertion (RPE) Scale.[19] Hutchinson and Tenenbaum[20] have been critical of this one-dimensional approach to the measurement of perceived exertion, and their findings support previous research[21-23] in demonstrating that several distinct inputs (sensory-discriminative, motivational-affective and cognitive-evaluative), perceived to different degrees during exertion, influence the perception of effort. Measurement of these variables has been effectively operationalised through direct questions during exertion. For example, Hutchinson and Tenenbaum[20] asked participants to rate their levels of muscular pain (sensory-discriminative), determination (motivational-affective) and effort (cognitive-evaluative) during physical exertion. The authors concluded that these sensations representing different dimensions of effort are perceived distinctly during exercise, and operate differently over the duration of an exertive task. Consequently, the validity of RPE and other single-item measures of effort, is questionable, and multidimensional measures that better represent the complex psychophysiological nature of effort are required. The potential lack of objectivity of psychological measures[24] may also be an important limitation to the use of RPE as a valid and reliable V̇O2max criterion. ObjecSports Med 2007; 37 (12) Maximal Oxygen Uptake Criteria tivity may be particularly compromised if researchers do not give appropriate pre-test instructions to subjects on how to interpret the RPE scale. Reduced pedal rate and power output have been used as objective measures to indicate that a subject has given a maximal effort. However, the reduced pedal rate or power output may be due to a lack of effort, rather than an indication that a maximum effort has been given. An observation that highlights the limitations of the currently used V̇O2max criteria is that we are unaware of any studies that have reported that one or more subjects did not satisfy the required V̇O2max criteria. It would be hard to believe that all subjects performing V̇O2max tests in these studies gave a maximum effort. Possible explanations are that the current criteria have insufficient sensitivity to detect subjects who did not give a maximal effort, or perhaps that some researchers have decided on appropriate criteria post hoc, to ensure no subjects have to be omitted from their study. The lack of standardisation of appropriate V̇O2max criteria certainly leaves the procedure of assessing the quality of a V̇O2max test susceptible to misuse. Although this brief critique has focused on the limitations of using criteria for assessing the quality of a V̇O2max test, it is noteworthy that this problem is relevant to all test protocols used in experimental research, where subjects are asked to push themselves to the limit of their exercise tolerance. Time trials and measures of time to exhaustion, in particular, are common in sport and exercise science research.[25,26] To the best of our knowledge, however, criteria have never been used in maximal tests other than V̇O2max tests, to establish whether a maximal effort has been given. This highlights an important area for future research. 3. Future Research Directions In view of the criticisms directed at currently used V̇O2max criteria,[5-7,10-12] it would be difficult to argue against the view that an improvement in the methodology of the existing criteria, or development of new criteria, is required. The remainder of this article discusses potential directions for future re 2007 Adis Data Information BV. All rights reserved. 1023 search that may eventually identify valid and robust V̇O2max criteria. 3.1 Verification Phase Thoden et al.[27] recommended that after the incremental phase of a V̇O2max test, athletes should rest for 5–15 minutes and then perform a constant speed run to exhaustion that is a speed equivalent to one stage higher than the last completed stage in the incremental phase. Thoden et al.[27] termed this latter procedure the ‘exhaustive phase’, and then later, the ‘verification phase’.[28] Although the verification phase was used only once in studies published in four journals surveyed over a 12-month period (table I), two recent studies involving sedentary individuals during cycling ergometry[29] and endurance athletes during treadmill running[11] reported that this procedure shows potential for yielding valid V̇O2max criteria. A V̇O2peak in the verification phase that is similar (within the tolerance of measurement error, e.g. within 2%[11]) to the V̇O2max value attained in the incremental phase would indicate a high probability that a true V̇O2max has been elicited. In addition to V̇O2max verification, Midgley et al.[11] suggested that HRmax verification would indicate that a subject has given a maximum effort, since it is improbable that a subject could give identical submaximal efforts during two exercise bouts with different characteristics (i.e. incremental exercise compared with a single square wave bout of exercise).[6] One of the main criticisms directed at current V̇O2max criteria is that they are specific to the exercise modalities, test protocols and subjects that were used in the studies where the criteria originated.[5,10-12] Since exercise modality, test protocol and subject characteristics will always vary between researchers and experimental studies, standardised V̇O2max criteria would need to be independent of these variables if they are to remain valid across studies. The V̇O2max and HRmax verification criteria appear to be independent of these variables because the criteria are not reliant on comparisons between the results of studies using different exercise modalities, test protocols and subjects. The test protocol Sports Med 2007; 37 (12) Midgley et al. 1024  2007 Adis Data Information BV. All rights reserved. VO2 (mL/min) . 3000 150 2000 100 1000 50 Heart rate (beats/min) 200 4000 0 0 0 100 200 300 400 500 600 1200 1300 1400 Time (sec) 200 4000 . 3000 150 2000 100 1000 50 Heart rate (beats/min) b VO2 (mL/min) 0 0 0 200 400 600 800 1000 Time (sec) 1800 200 3000 150 2000 100 1000 50 . Heart rate (beats/min) c 4000 0 0 0 20 0 40 0 60 0 80 0 10 0 12 0 0 14 0 0 16 0 00 24 0 26 0 00 A potential limitation of the verification phase is that due to accumulated fatigue during the incremental phase, for some subjects, the verification phase may be too short to allow V̇O2 and heart rate to reach their maximum.[11] Performing the incremental and verification phases on separate days[12,31] may increase time to exhaustion in the verification phase, by eliminating the residual fatigue effects of the preceding incremental phase when performed on the same day. Limitations to this approach are that an extra visit to the laboratory would be required and the day-to-day variation in V̇O2max[14] would reduce the robustness of the verification procedure. Thoden[28] recommended that the verification phase can be extended by performing it at the same, or a lower speed, than the peak speed attained in the incremental phase. This approach has been used in experimental studies by Day et al.[6] and Rossiter et al.[29] that both involved cycling ergometry and male subjects with a wide range of age and levels of fitness. However, this approach does not incorporate the principle of a V̇O2 plateau, in that V̇O2 demand has increased due to increased exercise intensity, but V̇O2 cannot increase sufficiently to meet this demand.[1] A multi-stage verification phase incorporating one or more relatively short stages of submaximal exercise, followed by a supramaximal effort may improve the verification procedure by allowing V̇O2 and heart rate more time to reach maximal values before volitional exhaustion occurs. Since the verification phase shows promise for yielding valid a . VO2 Heart rate VO2 (mL/min) independence of the verification criteria has been supported by a study that reported no statistically significant differences between maximal V̇O2 and heart rate values in the incremental and verification phases of three different V̇O2max test protocols that each incorporated a verification phase.[30] Figure 1 shows the V̇O2 and heart rate responses of a representative subject from this study during the incremental and verification phases of the three test protocols. Further research is required to establish the efficacy of the verification phase for yielding V̇O2max criteria that remain valid across studies using different experimental methodology and subjects. Time (sec) Fig. 1. Oxygen uptake (V̇O2) and heart rate responses of a 34-year-old, regional level, male long-distance runner during three maximal oxygen uptake (V̇O2max) tests, each incorporating different incremental test phases.[30] The incremental phases were: (a) a continuous protocol with 1-minute stage durations; (b) a discontinuous protocol with 2-minute stage durations and 30-second rest periods; and (c) a discontinuous protocol with 3-minute stage durations and 30-second rest periods. Each incremental phase was followed by a 10-minute rest phase and a verification phase. The verification phase consisted of a run to exhaustion at a speed that was equivalent to one stage higher than the last completed stage in the incremental phase. The similarities between the maximal V̇O2 and heart rate values in the incremental and verification phases confirms (verifies) that V̇O2max and the maximal heart rate were elicited during the increment phases. The horizontal dashed and dotted lines are to aid comparison of V̇O2max and heart rate values attained in the incremental and verification phases. Sports Med 2007; 37 (12) Maximal Oxygen Uptake Criteria V̇O2max criteria, future research should investigate different verification test protocols and criteria to improve the validity and utility of this procedure. 3.2 Oxygen Uptake Plateau The threshold value for the V̇O2 plateau criterion has typically been based on some arbitrary value, or the mean change in V̇O2 per stage increment for a research sample. The expected change in V̇O2 based on the American College of Sports Medicine metabolic equation (V̇O2 = [10.8 • W/m] + 7) has also been used to determine whether a V̇O2 plateau has occurred.[32] Although the validity of these approaches is questionable,[33] we believe that this does not exclude the V̇O2 plateau from being used as a valid V̇O2max criterion. However, research is required to develop new methodology for the identification of a valid and reliable V̇O2 plateau criterion. Since there is between-subject variation in the change in V̇O2 for any particular increase in exercise intensity,[4] the V̇O2 plateau criterion would be valid only if it was based on the individual slope of the V̇O2-exercise intensity relationship.[6] This approach also allows determination of the ‘total V̇O2 plateau’, which can be defined as the total deviation from linearity of the V̇O2-exercise intensity relationship. This is in contrast to currently used V̇O2 plateau criteria that are based on the difference in V̇O2 between only two stages or V̇O2 data points. The major limitation of using only two V̇O2 data points is that the determination of V̇O2 is associated with random errors of measurement, which during repeated measurements would scatter around the true value.[34] This ‘noise’ in V̇O2 data may result in the false identification of a V̇O2 plateau, and may be particularly problematic in pseudo-ramp protocols that are characterised by small increments in speed or power output. Furthermore, differences in V̇O2 between stages, as the basis for the identification of the V̇O2 plateau, is not feasible for ramp protocols. The V̇O2 plateau criterion could be based on the difference between the expected V̇O2 for the last stage of the test, calculated from the V̇O2-exercise intensity relationship and the actual value attained during the last stage. An appropriate V̇O2 plateau  2007 Adis Data Information BV. All rights reserved. 1025 threshold value is difficult to ascertain, but should be sufficiently stringent to indicate that the V̇O2 response has slowed enough to suggest that the individual’s rate of oxygen transport and utilisation was at or approaching its maximum. Future research should investigate the validity and reliability of a V̇O2 plateau criterion based on the individual slope of the V̇O2-work-rate relationship, as well as identifying an appropriate plateau threshold value. 3.3 Psychological Measures Based on current evidence, the validity of RPE and other single-item measures of perceived effort for evaluating the quality of a V̇O2max test, is questionable. Hutchinson and Tenenbaum[20] recommended the development and use of a more reliable and valid multidimensional assessment that more accurately captures perceptions of effort and fatigue and we further suggest that this approach should be considered for the development of new V̇O2max criteria. Such measures should take the form of visual analogue scales, which are easy to complete and have demonstrated good reliability and validity in clinical and health settings.[35] In order to establish whether participants have given a maximal effort and to indicate the limiting factors to further sustain a V̇O2max test, visual analogue scales can be used to assess aspects such as muscular pain, determination, and overall perceived effort.[20,36] For example, a V̇O2max criterion could be based on attaining a value of at least 8 on each visual analogue scale incorporating a range of 0–10. In addition to evaluating the relative level of effort a subject has given, a multidimensional approach may also provide an insight into why a subject has not given a maximal effort. This information may prove useful for developing strategies in the future that increase the likelihood that a subject will give a maximum effort during a V̇O2max test. Lack of objectivity may be viewed as a limitation to the development of psychological measures for use as valid and reliable V̇O2max criteria. Appropriate anchoring labels for scale values and verbal instructions by the researcher prior to testing[37] should help increase objectivity. Environmental and Sports Med 2007; 37 (12) 1026 personal aspects may influence psychological measures and environmental variables such as distracting stimuli and the presence and behaviour of others (including the researchers) would need to be standardised as much as possible during the development and use of psychology-based V̇O2max criteria. In addition to multi-dimensional approaches to measuring perceived effort during a V̇O2max test, we believe that a subject’s willingness and perceived ability to give a maximal effort should be assessed immediately prior to the test. Evidence suggests that psychological variables such as task-specific determination, perceived competence in exertion tolerance, willingness to invest effort, and physical selfefficacy greatly influence perseverance with physical exertion.[38] Not considering a subject’s psychological readiness to provide a maximum effort may therefore be an important omission. Measurement of psychological readiness to give a maximal effort during a V̇O2max test could easily be achieved. Tenenbaum et al.[38] reported an effective and simple approach for strength and endurance tasks, with pretask questions of “how confident are you in tolerating this exertion and discomfort?” (specific selfefficacy), “how determined and committed are you to perform the task?” (degree of commitment) and “how much effort do you intend to put into the task?” (effort investment). This ‘auxiliary’ information could be used to better interpret formal V̇O2max criteria. For example, a low score on “how much effort do you intend to put into the task?” may help support V̇O2max criteria that indicate a subject has given a poor effort. Additionally, such psychological readiness profiling may allow a researcher to make an informed judgement whether to continue with a V̇O2max test that may result in invalid physiological data, due to poor subject motivation, for example. Psychological readiness profiling for maximal testing could be an interesting area for future research. 4. Conclusions The currently used V̇O2max criteria have been heavily criticised in relation to their validity and applicability across studies. Additionally, previous  2007 Adis Data Information BV. All rights reserved. Midgley et al. studies have typically used several criteria, but there has been no agreement on how many criteria should be used, or the proportion that need to be satisfied to confirm the validity of the V̇O2max test results. With this lack of standardisation, researchers may be tempted to decide on criteria post hoc, to ensure that no subjects have to be omitted from their study. The above limitations considerably reduce the confidence that the subjects in an experimental study elicited a ‘true’ V̇O2max, or that a maximum effort was given. The credibility of V̇O2max as a robust measure of cardiorespiratory fitness could therefore be questioned, particularly for cross-sectional studies where comparisons between studies may be affected by differences in experimental methods, or the subjects under investigation. This point is also relevant for performance measures such as the peak power output[39] or treadmill speed,[40] often associated with V̇O2max tests. In view of the above limitations, we believe that future research should attempt to identify a new set of standardised V̇O2max criteria. In order to be universally applied, each criterion needs to be independent of exercise modality, test protocol and subject characteristics. In this regard, the verification procedure and the V̇O2 plateau criterion based on the individual slope of the V̇O2-work-rate relationship have shown promise. We also recommend a multidimensional psychophysiological approach to evaluating the relative degree of effort, to replace the one-dimensional measure of the currently used RPE scale. Psychological measures that assess the subject’s willingness to give a maximum effort before the test might also provide useful ‘auxiliary’ information, particularly in relation to a subject’s readiness for maximal testing. A major consideration for the development of a new set of standardised V̇O2max criteria is how to interpret the criteria when they have not all been satisfied. In fact, how well all V̇O2max criteria agree with each other may be a good indication of the specificity and sensitivity[12] of the criteria in detecting whether or not an individual has elicited V̇O2max or given a maximum effort. If all V̇O2max criteria demonstrate a high degree of specificity and Sports Med 2007; 37 (12) Maximal Oxygen Uptake Criteria sensitivity, they should either all be satisfied, or all not satisfied. Howley et al.[5] suggested that published guidelines are required to promote uniformity in V̇O2max testing procedures. We further suggest that these guidelines, such as an American College of Sports Medicine Position Stand, should recommend a standardised set of V̇O2max criteria. At present, however, the greatest challenge is identifying V̇O2max criteria that remain valid across studies that are using different test methodology and subjects. Acknowledgements No sources of funding were used to assist in the preparation of this article. The authors have no conflicts of interest that are directly relevant to the content of this article. References 1. Hill AV, Lupton H. Muscular exercise, lactic acid, and the supply and utilization of oxygen. Q J Med 1923; 16: 135-71 2. Buchfuhrer MJ, Hansen JE, Robinson TE, et al. Optimizing the exercise protocol for cardiopulmonary assessment. J Appl Physiol 1983; 55: 1558-64 3. Robergs RA, Burnett AF. Methods used to process data from indirect calorimetry and their application to V̇O2max. J Exerc Physiol 2003; 6: 44-57 4. Taylor HL, Buskirk E, Henschel A. Maximal oxygen intake as an objective measure of cardio-respiratory performance. J Appl Physiol 1955; 8: 73-80 5. Howley ET, Bassett Jr DR, Welch HG. Criteria for maximal oxygen uptake: review and commentary. Med Sci Sports Exerc 1995; 27: 1292-301 6. Day JR, Rossiter HB, Coats EM, et al. The maximally attainable V̇O2 during exercise in humans: the peak vs maximum issue. J Appl Physiol 2003; 95: 1901-7 7. Duncan GE, Howley ET, Johnson BN. Applicability of V̇O2max criteria: discontinuous versus continuous protocols. Med Sci Sports Exerc 1997; 29: 273-8 8. Astrand PO. Experimental studies of physical working capacity in relation to sex and age. Copenhagen: Ejnar Munksgaard, 1952 9. Issekutz B, Birkhead NC, Rodahl K. Use of the respiratory quotients in assessment of aerobic work capacity. J Appl Physiol 1962; 17: 47-50 10. Cumming GR, Borysyk LM. Criteria for maximum oxygen uptake in men over 40 in a population survey. Med Sci Sports 1972; 4: 18-22 11. Midgley AW, McNaughton LR, Carroll S. Verification phase as a useful tool in the determination of the maximal oxygen uptake of runners. Appl Physiol Nutr Metab 2006; 31: 541-8 12. Stachenfeld NS, Eskenazi M, Gleim GW, et al. Predictive accuracy of criteria used to assess maximal oxygen consumption. Am Heart J 1992; 123: 922-5 13. Meyer T, Scharhag J, Kindermann W. Peak oxygen uptake: myth and truth about an internationally accepted reference value. Z Kardiol 2005; 94: 255-64  2007 Adis Data Information BV. All rights reserved. 1027 14. Katch VL, Sady SS, Freedson P. Biological variability in maximum aerobic power. Med Sci Sports Exerc 1982; 14: 21-5 15. Londeree BR. Influence of age and other factors on maximal heart rate. J Cardiac Rehabil 1984; 4: 44-9 16. Medbo JI, Sejersted OM. Acid-base and electrolyte balance after exhausting exercise in endurance-trained and sprinttrained subjects. Acta Physiol Scand 1985; 125: 97-109 17. Meyer T, Welter JP, Scharhag J, et al. Maximal oxygen uptake during field running does not exceed that measured during treadmill exercise. Eur J Appl Physiol 2003; 88: 387-9 18. Hoffman RL. Effect of training at the ventilatory threshold on the ventilatory threshold and performance in trained distance runners. J Strength Cond Res 1999; 13: 118-23 19. Borg GA. Perceived exertion. Exerc Sport Sci Rev 1974; 2: 131-53 20. Hutchinson JC, Tenenbaum G. Perceived effort: can it be considered gestalt? Psychol Sport Exerc 2006; 7: 463-76 21. Hardy CJ, Rejeski WJ. Not what, but how one feels: the measurement of affect during exercise. J Sport Exerc Psychol 1989; 11: 304-17 22. McAuley E, Courneya KS. The subjective exercise experience scale (SEES): development and preliminary validation. J Sport Exerc Psychol 1994; 16: 163-77 23. Parfitt G, Markland D, Holmes C. Responses to physical exertion and psychological affect in the early stages of exercise. Percept Mot Skills 1994; 80: 259-66 24. Rosenberg S. Objectivity in psychology and philosophy. Am Psychol 1980; 35: 219-22 25. Burgomaster KA, Heigenhauser GJ, Gibala MJ. Effect of shortterm sprint interval training on human skeletal muscle carbohydrate metabolism during exercise and time-trial performance. J Appl Physiol 2006; 100: 2041-7 26. Basset FA, Chouinard R, Boulay MR. Training profile counts for time-to-exhaustion performance. Can J Appl Physiol 2003; 28: 654-66 27. Thoden JS, MacDougall JD, Wilson BA. Testing aerobic power. Ithaca (NY): Mouvement Publications Inc., 1982 28. Thoden JS. Testing aerobic power. In: MacDougall JD, Wenger HA, Green HJ, editors. Physiological testing of the highperformance athlete. 2nd ed. Champaign (IL): Human Kinetics, 1991: 107-73 29. Rossiter HB, Kowalchuk JM, Whipp BJ. A test to establish maximum O2 uptake despite no plateau in the O2 uptake response to ramp incremental exercise. J Appl Physiol 2006; 100: 764-70 30. Midgley AW, McNaughton LR, Carroll S. Time at V̇O2max during intermittent treadmill running: test protocol dependent or methodological artefact? Int J Sports Med. Epub 2007 May 11 31. Morgan DW, Baldini FD, Martin PE, et al. Ten kilometer performance and predicted velocity at V̇O2max among welltrained male runners. Med Sci Sports Exerc 1989; 21: 78-83 32. Wingo JE, Lafrenz AJ, Ganio MS, et al. Cardiovascular drift is related to reduced maximal oxygen uptake during heat stress. Med Sci Sports Exerc 2005; 37: 248-55 33. Saitoh M, Matsunaga A, Kamiya K, et al. Comparison of cardiovascular responses between upright and recumbent cycle ergometers in healthy young volunteers performing low-intensity exercise: assessment of reliability of the oxygen uptake calculated by using the ACSM metabolic equation. Arch Phys Med Rehabil 2005; 86: 1024-9 Sports Med 2007; 37 (12) Midgley et al. 1028 34. Wyndham CH, Strydom NB, Maritz JS, et al. Maximum oxygen intake and maximum heart rate during strenuous work. J Appl Physiol 1959; 14: 927-36 35. Nyenhuis DL, Stern RA, Yamamoto C, et al. Standardization and validation of the visual analogue mood scales. Clin Neuropsychol 1997; 11: 407-15 36. Ueda T, Nabetani T, Teramoto K. Differential perceived exertion measured using a new visual analogue scale during pedaling and running. J Physiol Anthropol 2006; 25: 171-7 37. Noble BJ, Robertson RJ. Perceived exertion. Champaign (IL): Human Kinetics, 1996 38. Tenenbaum G, Hall HK, Calcagnini N, et al. Coping with physical exertion and negative feedback under competitive and self-standard conditions. J Appl Soc Psychol 2001; 31: 1582-626  2007 Adis Data Information BV. All rights reserved. 39. Schabort EJ, Killian SC, St Clair Gibson A, et al. Prediction of triathlon race time from laboratory testing in national triathletes. Med Sci Sports Exerc 2000; 32: 844-9 40. Noakes TD, Myburgh KH, Schall R. Peak treadmill running velocity during the V̇O2max test predicts running performance. J Sports Sci 1990; 8: 35-45 Correspondence: Dr Adrian W. Midgley, Department of Sport, Health and Exercise Science, University of Hull, Cottingham Road, Hull, HU6 7RX, UK. E-mail: A.W.Midgley@hull.ac.uk Sports Med 2007; 37 (12)