Journal Logo

Special Communications: Methods

Applicability of ˙VO2max criteria: discontinuous versus continuous protocols


Author Information
Medicine & Science in Sports & Exercise: February 1997 - Volume 29 - Issue 2 - p 273-278
  • Free


One of the most common measurements in exercise physiology is maximal oxygen uptake (˙VO2max), which describes an individual's capabilities for the uptake, transport, and utilization of oxygen(15). Obtaining valid and reliable ˙VO2max values is of considerable physiological importance, particularly when comparing individuals or groups, when following subjects longitudinally, and when different modes of exercise are used (11). To accomplish this it is necessary to establish a series of standards, that once achieved, will increase the likelihood that a true ˙VO2max value has been obtained. Traditionally, a plateau in ˙VO2 has been viewed as the best objective criterion for establishing ˙VO2max(21). However, a ˙VO2 plateau is not always attained in all subjects tested(2,4,5,17,22). As described by Cumming and Borysyk (4), in the absence of a plateau other criteria have been used to indicate that a ˙VO2max value has been achieved, including an increase in heart rate to maximum values estimated for age (13), a respiratory exchange ratio (RER) of 1.15 or greater (9), and high post exercise blood lactate levels, usually above 8 mmol·l-1(2). All of these standards were derived from discontinuous, often multiday, protocols. Although there is good evidence that the ˙VO2max value itself is not different when measured by a continuous versus a discontinuous protocol(12,14), there is some question about the applicability of the above criteria to continuous protocols(8).

The purpose of this study was to compare the classic discontinuous Taylor et al. (21) protocol to a continuous version of the same protocol to evaluate (a) maximal physiological responses, and (b) the quantitative values and incidence of achievement of the various˙VO2max criteria. Finally, comments will be offered on the importance of using the ˙VO2 plateau criterion as the primary standard defining the attainment of ˙VO2max and how we might use the secondary criteria to evaluate the quality of a graded exercise test.


Ten healthy male volunteers gave their signed informed consent to participate in this investigation. Power analysis, using 10 subjects and the RER and lactate variables at an alpha level of 0.05, indicated values of 0.84 and 0.75, respectively (3). None of the subjects were taking any medications that might interfere with physiological responses, and none had any limitations to strenuous exercise as determined by a health history questionnaire. Subjects reported that they were active (i.e., running, bicycling, weight lifting, and rock climbing approximately 3-4 times per week), however, none were engaged in regular structured training programs. All subjects reported that they had completed a maximal graded exercise test within the past 2-3 yr. This study consisted of three separate tests: an orientation and ˙VO2max prediction trial; a multiday, discontinuous treadmill protocol (DT); and, a continuous treadmill protocol (CT). The orientation trial was always first; however, the order of the DT and CT trials was counterbalanced. Each test was separated by a period of 2-3 days. Subjects were asked to refrain from eating for a minimum of 2 h prior to each test and from vigorous exercise on the day of and the day before the test. All testing was performed in an air-conditioned laboratory maintained at approximately 22°C.

Age, height, weight, and skinfold thicknesses were obtained during the orientation trial. Weight was measured on a calibrated physicians scale and height was measured on a wall-mounted stadiometer. Skinfolds were taken with a Lange caliper on the right side of the body and included the chest, abdomen, and thigh. Percent fat was determined according to the formula of Pollock et al. (18). All subjects were instructed on the use of the Borg 6-20 Rating of Perceived Exertion (RPE) scale to ascertain their perception of effort during each test. Subjects completed a ˙VO2max prediction test on a calibrated motor-driven treadmill (Quinton Corp., Seattle, WA), which served as a practice session to estimate˙VO2max(20,21). This test involved level running at 6.5, 7.0, and 7.5 mph for 3 min at each level. Heart rate was obtained with a heart rate monitor (Polar Electro, Kempele, Finland) and recorded at the end of every stage. Likewise, RPE was obtained at the end of every stage. To predict ˙VO2max and establish the first workload for the DT, heart rate was plotted against the predicted ˙VO2(1) for each level running speed on the prediction test and extrapolated to age predicted maximal heart rate (220-age). The predicted˙VO2max was determined from the straight line regression equation between oxygen cost at each running speed and heart rate.

The DT consisted of a series of 2-4 workloads performed at 7.0 mph and 2.5% increases in treadmill elevation. The different stages of the DT were separated by 2-3 d. For each workload, a 10-min warm-up at 6.0 mph and 0% grade was performed, followed by a 5-min rest period. Following this, the subjects ran at 7.0 mph and the grade estimated to elicit ˙VO2max(1) as determined from the prediction trial. Subjects ran for a minimum of 3 and a maximum of 5 min at each grade. Oxygen uptake was measured continuously using an on-line metabolic program. In addition, gas samples were collected in Douglas bags from a port at the bottom of the mixing chamber which allowed us to verify the gas fractions obtained from the on-line system with those collected in the bags. After a 2.5-min period of walking at 3.0 mph and 0% grade, subjects were seated to collect blood for lactate analysis. A finger stick blood sample (50 ul) was collected at 3, 4, and 5 min post-exercise. Each sample was deproteinized in 2 ml of cold perchloric acid, centrifuged, and stored frozen for future analysis. The samples were analyzed in duplicate for lactate by the enzymatic flourometric analysis technique(7). The highest value was used for comparison purposes without reference to any particular time point. This entire procedure was repeated on the next testing date at an elevation that was 2.5% higher than the first. Again, subjects ran for a period of 3-5 min. This continued, each stage 2.5% higher than the previous, until the subject could no longer complete a minimum of 3 min at a given workload. It should be noted that the DT was continued even if a plateau in ˙VO2 had been observed after only two workloads. Therefore, the endpoint of the DT was an inability of the subject to complete 3 min at a given workload, and not the demonstration of an apparent plateau. Subjects received strong verbal encouragement throughout the entire test; however, the test was terminated when the subject could no longer continue and signaled to stop. Heart rate and RPE were recorded at the end of each stage.

For the CT, after an identical warm-up and rest period as used for the DT, subjects ran at 7.0 mph and 0% grade for 3 min. Gas samples were collected continuously as described previously. Following this, treadmill elevation was increased 2.5% every minute until the subject could no longer continue the test. In every instance, subjects were encouraged to finish the entire 1-min stage. If they could not complete a stage, gas samples from the partially completed stage were not used in data analysis due to the nature of the gas sampling interval. Only two subjects were not able to finish a final 1-min stage on the CT (average time in a new stage was 10 s) and all completed a final 3-min stage on the DT. Blood was collected at the end of the test in a manner previously described. Heart rate was recorded every minute, and RPE was recorded at the end of every stage starting at the end of minute 3.

During the tests, subjects breathed room air through a Hans-Rudolph (Kansas City, MO) two-way, non-rebreathing valve (2700 series). Volume(˙VI) was measured during inspiration using a Rayfield ventilation meter (Rayfield Equipment Ltd., Waitsfield, VT). The calibration of this instrument was verified prior to the start of this study with a Collins 120 L spirometer. Gas was sampled from a 5-L mixing chamber by an Applied Electrochemistry (Ametek Corp., Pittsburgh, PA) S-3A/1 O2 analyzer and an Applied Electrochemistry CD-3A CO2 analyzer, respectively. Prior to testing, the analyzers were calibrated with room air and two gases of known concentration. The composition of these gases was verified by micro-Scholander technique (19). Therefore, a 3-point calibration scheme was employed ensuring linearity of the gas analyzers. Pulmonary gas exchange was measured continuously and metabolic variables were calculated utilizing a 1-min data averaging technique by a Rayfield metabolic program (Rayfield Equipment Ltd.). Because error may be introduced by the use of automated systems if there is not a matching of gas fractions and ventilation, particularly when both are changing at a high rate during maximal work(8), Douglas bags were also collected at 1-min intervals to coincide with the on-line system. After the test was completed, the analyzers were again calibrated as described previously to analyze the gas fractions in the bags.

Data are reported as [horizontal bar over]X ± SD. Maximal physiological values between protocols were compared using a t-test for dependent samples. Data from the Rayfield metabolic program were used for comparison purposes, while values from the Douglas bags were used as a quality control. Specific criteria used in this investigation were a plateau in˙VO2 (change in ˙VO2 < 150 ml·min1, or 2.1 ml·kg-1·min-1 between successive treadmill elevations) (21), a heart rate equal to age adjusted maximal values (13), an RER ≥ 1.15(9), and blood lactate levels > 8 mmol·l-1(2). A chi-square test was used to determine if there were any statistical differences in the frequency of criteria achievement between protocols. Statistical significance was set atP < 0.05.


The age, height, weight, and body composition of the ten male subjects were 24.1 ± 2.5 yr, 176.6 ± 6.9 cm, 75.7 ± 11.5 kg, and 9.8± 4.2% fat, respectively. The maximal physiological values obtained on each protocol are presented in Table 1. Absolute and weight relative ˙VO2max were similar (P > 0.05) between protocols. The accompanying values from the Douglas bags were 4.2± 0.5 vs 4.2 ± 0.4 l·min-1 and 55.7 ± 3.4 vs 55.7 ± 3.5 ml·kg-1·min-1 for the DT and the CT, respectively. Likewise, there were no differences (P > 0.05) with respect to ˙VI, RPE, and HR between protocols. Maximal RER and blood lactate were greater (P < 0.05) during the DT as compared to the CT. The RER values from the Douglas bags were within 0.02 of those from the Rayfield system.

The results for the achievement of the various ˙VO2max criteria appear in Figure 1. Only 10% of the subjects on the DT, and 40% on the CT, achieved the heart rate standard. Applying ± 1 SD to our subjects increased the achievement of the heart rate standard to 70% on the CT and 60% on the DT. A plateau was achieved by 60% of the subjects on the DT and 50% of the subjects on the CT. The RER and blood lactate criteria were achieved by 100% of the subjects on the DT and 90% of the subjects on the CT. The chi-square test revealed that the row (protocol) by column (criteria) variables were independent of each other (P > 0.05), indicating that criteria achievement was independent of test protocol.


The first aim of this study was to compare the maximal physiological responses obtained during continuous and discontinuous graded treadmill protocols. In agreement with other reports(12,14), we found that ˙VO2max was similar during each of the two protocols. The finding that ˙VI was similar between tests is in agreement with Maksud and Coutts(12). In contrast to their findings, we found that heart rate was similar between protocols. However, there was a trend for heart rate to be higher during the CT as compared to the DT (P = 0.06). This is probably a reflection of the longer running time experienced during the CT since subjects generally achieved the same treadmill elevation at maximal exercise for each protocol. Maximal values for RER and blood lactate were significantly greater for the DT as compared to the CT. This is reasonable since exercise performed at a supra-maximal workload would theoretically result in an increased oxygen deficit, hence, requiring a greater anaerobic energy requirement to supply ATP at the onset of exercise. As a result, blood lactate concentration would be higher, thus increasing RER owing to the increased buffering of lactate (10).

The second aim of this study was to compare the quantitative values and incidence of achievement of the various ˙VO2max criteria between protocols. The age-adjusted estimate of maximal heart rate standard was the most difficult to achieve, with 40% of the subjects on the CT and 10% on the DT satisfying this criterion. While achievement of an absolute age-adjusted maximum value for heart rate is a practical criterion for having achieved maximum effort, it may also be the most unreliable because of the individual variability and large standard deviation (± 11 beats·min-1) associated with an estimate of maximum heart rate(4). Quantitatively, the maximal values achieved for heart rate on each protocol fell well below an absolute age-adjusted standard(approximately 196 for this population). Applying ± 1 SD to our subjects increased the achievement of the heart rate criterion to 70% on the CT and 60% on the DT. As previously mentioned, the variability in maximal heart rate is too large to use average maximum heart rate alone as a criterion for ˙VO2max(4).

To determine the magnitude and limits of the change in ˙VO2 as a result of increasing treadmill grade by 2.5% before ˙VO2max is attained, Taylor et al. (21) measured ˙VO2 in 13 of their subjects at 2 or more grades below the grades resulting in a plateau and found that the mean ˙VO2 increment for their testing procedures was 299.3 ml·min-1, or 4.18 ml·kg-1·min-1. Therefore, a plateau in˙VO2 was considered to occur if two successive stages, separated by an elevation of 2.5%, differed by less than 150 ml·min1, or 2.1 ml·kg-1·min-1. We found that for our procedures, the mean ˙VO2 increment between stages was 286 ml·min-1 and 3.9 ml·kg-1·min-1. Therefore, we were able to apply with confidence the plateau criterion proposed by Taylor et al. (21). By applying this cut-off value, 50% and 60% of subjects met the criteria established for a˙VO2 plateau on the CT and DT, respectively. However, plateau achievement would have increased to 80% on the DT if the procedures used by Taylor et al. (21) were followed strictly. For example, there were two instances when a plateau had been observed between the first and second workloads. However, completion of a third workload, 2.5% higher than the previous, resulted in a change in ˙VO2 beyond that specified for a plateau. This is consistent with the findings of Glassford et al. (6), who found that the use of an arbitrary cut-off value for plateau determination can result in an underestimation of˙VO2max.

The RER and blood lactate criteria were the most easily achieved. It is assumed that the rise in RER during heavy exercise is due to an imbalance between the formation and elimination of acids, particularly lactic acid, and to their reaction with the bicarbonate pool of the body(10). Therefore, it seems logical that if the lactate criterion were achieved, then the RER criterion would also be achieved. The incidence of achievement of these two criteria were the same for both protocols; 90% for the CT and 100% for the DT. Quantitatively, the mean values achieved on each protocol were well beyond the established 1.15 RER and 8 mmol·l-1 blood lactate standards. This finding would indicate that both of these classic criteria, although established from discontinuous protocols, are applicable to continuous protocols. According to Astrand(2) and Issekutz et al. (9), the achievement of these criteria would indicate that an individual was working at or near maximal effort.

The present data are in agreement with others(2,4-6,16,20) who have suggested that many subjects stressed to maximal effort will not demonstrate a plateau. This is not to say that these subjects have not achieved their“true” ˙VO2max. Figure 2A depicts the plot of a typical subject who demonstrated a plateau on both protocols. Note that the linear increase in ˙VO2 during the CT was interrupted by the final stage only. In addition to the demonstration of a plateau, this subject also satisfied the RER and lactate criteria on both tests. Therefore, it would appear that this subject had in fact provided a maximal effort and˙VO2max was obtained. Contrast this figure with that ofFigure 2B, which depicts the plot of a subject who did not plateau on either protocol. Note that there was essentially no departure from linearity during either test. While this subject did not demonstrate a plateau, he did achieve the RER, heart rate, and lactate criteria on the CT, and the RER and lactate criteria on the DT. Is it reasonable to assume that this subject did not achieve ˙VO2max just because he did not plateau? Achievement of the other criteria suggests that he did perform to maximal effort. Herein lies the dilemma. How should the highest˙VO2 value obtained during a graded exercise test be interpreted in the absence of a plateau?

The consistent finding that ˙VO2max is similar between a continuous protocol, where plateau achievement is 50% or less, and a discontinuous protocol, where plateau achievement is approximately 60-90%, would suggest that there is little to be gained in using a ˙VO2 plateau as the primary criterion for having achieved ˙VO2max. Therefore, researchers should use the secondary objective criteria to validate the ˙VO2max values obtained. However, as noted by Shephard(20), it is not clear how to use these secondary criteria, alone or in combination, to evaluate the quality of a graded exercise test when a true plateau in ˙VO2 does not occur.

Despite the problems associated with the heart rate standard, Howley et al.(8) found that 9 of 29 studies published inMedicine and Science in Sports and Exercise (Oct. 1993-May 1994) cited the attainment of some percentage of age-adjusted maximal heart rate as a criterion for defining ˙VO2max. The heart rate standards cited include ± 10 beats·min-1 of (220-age), ± 15 beats·min-1 of (220-age), heart rate ≥ (220-age), and heart rate ≥ 90% of (220-age). It is not clear how such “standards” have evolved into the literature. Cumming and Borysyk (4) concluded more than 20 years ago that the maximal heart rate range was too wide to use an average maximum value alone as a criterion for˙VO2max, a conclusion that is as reasonable today as it was then. Thus, out of the four “classic” criteria, only the RER and blood lactate standards appear to be of general value. Unfortunately, Howley et al.(8) found that only one in 29 of the studies mentioned previously used the lactic acid standard, while 7 in 29 studies failed to report the achievement of any criteria at all. Despite the significance of the measurement of ˙VO2max, to date there is no general agreement on the use of specific criteria for establishing that a ˙VO2max value has been achieved. The use of subjective criteria such as RPE and investigator evaluation may be additional ways by which to assess the quality of a graded exercise test.

Based on the results of this investigation, the following conclusions can be applied to young fit males. Continuous treadmill protocols result in maximal physiological values, with the exception of RER and blood lactate, that are similar to those obtained during discontinuous protocols. The achievement of the various ˙VO2max criteria are independent of test protocol. Therefore, although the currently used objective criteria were developed from discontinuous protocols, it appears that they are applicable to continuous protocols. A ˙VO2 plateau is not an absolute prerequisite for defining an individual's ˙VO2max and is of limited use as the primary objective criterion for evaluating the quality of a graded exercise test. Therefore, the achievement of secondary objective criteria, specifically RER and lactate in combination, increases the likelihood that the highest ˙VO2 value achieved is ˙VO2max.

Figure 1-Proportion of subjects achieving ˙VO2max criteria during discontinuous (DT) and continuous (CT) treadmill protocols.
Figure 1-Proportion of subjects achieving ˙VO2max criteria during discontinuous (DT) and continuous (CT) treadmill protocols.
Figure 2-A) Graphical representation of a subject demonstrating a˙VO2 plateau on the discontinuous (DT) and continuous (CT) protocol. B) Graphical representation of a subject failing to demonstrate a˙VO2 plateau on either the discontinuous (DT) or continuous (CT) protocol.
Figure 2-A) Graphical representation of a subject demonstrating a˙VO2 plateau on the discontinuous (DT) and continuous (CT) protocol. B) Graphical representation of a subject failing to demonstrate a˙VO2 plateau on either the discontinuous (DT) or continuous (CT) protocol.


1. American College Of Sports Medicine. Guidelines for Exercise Testing and Prescription, 4th Ed. Philadelphia: Lea & Febiger, 1991, pp. 296-299.
2. Åstrand, P.-O. Experimental Studies of Physical Working Capacity in Relation to Sex and Age. Copenhagen: Ejnar Munksgaard, 1952, pp. 23-27, 92-102.
3. Cohen, J. Statistical Power Analysis for the Behavioral Sciences (Revised Ed.). New York: Academic Press, 1977, pp. 19-52.
4. Cumming, G. R. and L. M. Borysyk. Criteria for maximum oxygen uptake in men over 40 in a population survey. Med. Sci. Sports Exerc. 4:18-22, 1972.
5. Froelicher, V. F., H. Brammell, G. Davis, I. Noguera, A. Steward, and M. C. Lancaster. A comparison of three maximal treadmill exercise protocols. J. Appl. Physiol. 36:720-725, 1974.
6. Glassford, R. G., G. H. Y. Baycroft, A. W. Sedgwick, and R. B. J. MacNab. Comparison of maximal oxygen uptake values determined by predicted and actual methods. J. Appl. Physiol. 20:509-513, 1965.
7. Gutmann, P. and A. Wahlefeld. L- (+) -lactate: determination with lactate dehydrogenase and NAD. In: Methods of Enzymatic Analysis, H. Bergmeyer (Ed.). New York: Academic Press, 1974, pp. 1464-1468.
8. Howley, E. T., D. R. Bassett, Jr., and H. G. Welch. Criteria for maximal oxygen uptake: review and commentary. Med. Sci. Sports Exerc. 27:1292-1301, 1995.
9. Issekutz, B., N. C. Birkhead, and K. Rodahl. The use of respiratory quotients in assessment of aerobic work capacity. J. Appl. Physiol. 17:47-50, 1962.
10. Issekutz, B. J. and K. Rodahl. Respiratory quotient during exercise. J. Appl. Physiol. 16:606-610, 1961.
11. Kasch, F. W., J. P. Wallace, R. R. Huhn, L. A. Krogh, and P. M. Hurl. ˙VO2max during horizontal and inclined treadmill running. J. Appl. Physiol. 40:982-983, 1976.
12. Maksud, M. G. and K. D. Coutts. Comparison of a continuous and discontinuous graded treadmill test for maximal oxygen uptake.Med. Sci. Sports Exerc. 3:63-65, 1971.
13. Maritz, J. S., J. F. Morrison, J. Peter, N. B. Strydom, and C. H. Wyndham. A practical method of estimating an individual's maximum oxygen intake. Ergonomics 4:97-122, 1961.
14. McArdle, W. D., F. I. Katch, and G. S. Pechar. Comparison of continuous and discontinuous treadmill and bicycle tests for max˙VO2Med. Sci. Sports. Exerc. 5:156-160, 1973.
15. McConnell, T. R. Practical considerations in the testing of ˙VO2max in runners. Sports Med. 5:57-68, 1988.
16. Noakes, T. D. Implications of exercise testing for prediction of athletic performance: a contemporary perspective. Med. Sci. Sports Exerc. 20:319-330, 1988.
17. Pollock, M. L., R. L. Bohannon, K. H. Cooper, et al. A comparative analysis of four protocols for maximal treadmill stress testing.Am. Heart J. 92:39-46, 1976.
18. Pollock, M. L., D. H. Schmidt, and A. S. Jackson. Measurement of cardiorespiratory fitness and body composition in a clinical setting. Comp. Ther. 6:12-27, 1980.
19. Scholander, P. Analyzer for accurate estimation of respiratory gases in one-half cubic centimeter samples. J. Biol. Chem. 167:235-250, 1947.
20. Shephard, R. J. Tests of maximum oxygen intake: a critical review. Sports Med. 1:99-124, 1984.
21. Taylor, H. L., E. Buskirk, and A. Henschel. Maximal oxygen intake as an objective measure of cardio-respiratory performance.J. Appl. Physiol. 8:73-80, 1955.
22. Taylor, H. L., Y. Wang, L. Rowell, and G. Blomqvist. The standardization and interpretation of submaximal and maximal tests of working capacity. Pediatrics 32:703-722, 1963.


©1997The American College of Sports Medicine