Maximal oxygen uptake (V˙O2max) is the main indicator of cardiorespiratory fitness (CRF). A low CRF level is a superior independent risk factor of all-cause mortality when compared with obesity, weight gain, smoking, hypertension, high cholesterol, and diabetes (3,9,18). Measuring V˙O2max from a maximal graded exercise test (GXT) is expensive, requires specialized equipment, and may be impractical for unfit or diseased populations. Furthermore, a GXT requires a physician present if the patient is at risk (1). Assessing CRF in the general public via submaximal GXT is essential for determining CRF.
The YMCA cycle ergometer test is a commonly used submaximal test for predicting V˙O2max. The test predicts V˙O2max by extrapolating heart rate (HR) responses relative to power output increases, whereby age predicted maximal heart rate (APMHR) is used to define maximal power output (12). Peak power output (Wpeak) is then used to estimate V˙O2max using the following metabolic equation (1).
where V˙O2max is expressed in milliliters per kilogram per minute, body mass (BM) is in kilograms, and Wpeak is in watts.
Predicting V˙O2max using the YMCA test has limitations. First, the APMHR equation (220 − age) recommended to predict V˙O2max has questionable scientific merit, with a standard error of estimate ranging from 6 to 22 bpm across studies (30). The only cross validation study to our knowledge validating the YMCA test used 220 − age (2,12,30). Second, the YMCA test uses a low cadence (i.e., 50 rpm), which may not reflect the optimum mechanical efficiency for every subject. Lower cadences yield higher gross efficiencies for untrained and noncyclists; however, Lucia et al. (20) reported gross efficiency and preferred cadence are higher among trained cyclists (e.g., 90–100 rpm) (10,13,20,27,29). Third, the YMCA test uses predetermined power output and test termination criteria that may not enable younger and fitter subjects to surpass their gas exchange threshold (GET) (8,12). Last, the YMCA test may omit a stage above GET failing to account for the nonlinearity of oxygen uptake associated with the V˙O2 slow component (19,34). These limitations emphasize the need for a submaximal test that improves the validity of estimating V˙O2max.
Use of demographic data and exercise domain responses from previously published research may increase the validity of submaximal testing for determining V˙O2max. Prescribing individualized exercise intensities in the moderate and heavy domains of exercise, using a nonexercising equation validated by Jackson et al. (15), may increase the accuracy for estimating V˙O2max via submaximal testing. Previously published normative threshold data by Davis et al. (5,6) makes it feasible to prescribe exercises intensity stages above and below estimated GET. Therefore, the purpose of this study was to develop and validate a customized submaximal cycle ergometer test that takes into account the nonlinear oxygen uptake-power output response above GET and yield a more accurate estimation of V˙O2max.
The experimental design included two separate visits to the laboratory. Before testing, subjects were assessed using ACSM Risk Stratification guidelines for maximal exercising testing (1). Subjects between the ages of 18 and 39 were included in the study if stratified as moderate risk or lower. Subjects 40 yr of age or older were included in the study if stratified as low risk. Visit 1 included risk stratification, informed consent, self-reported physical activity rating (PA-R), measurements of subject’s height and body mass, and administering a GXT with square wave verification bout (17). Visit 2 included two submaximal exercise tests: our new Mankato submaximal exercise test (MSET) and the YMCA cycle ergometer test (note: We chose to name the test after our affiliation to avoid confusion with the 6-min walk test). Each submaximal test was counterbalanced to avoid an order effect. There was 30-min rest period between the two submaximal tests. Before each exercise test, a 5-min warm-up was administered at 50 W, followed by a 5-min passive rest. Preferred cadence for the GXT and MSET was determined during the initial warm-up session.
A total of 56 subjects were included in the study (31 men and 26 women; age, 28.4 ± 10.4 yr; BMI, 24.4 ± 3.2). Written informed consent was obtained before the start of the study. All subjects self-reported abstaining from the use of alcohol or engaging in heavy exercise 24 h preceding each visit. Subjects were given at least 48 h of rest between visits, and all tests were completed within 2 wk. Our institutional review board for the protection of human subjects approved all procedures for this study.
All tests were conducted using friction-based, weight-loaded cycle ergometer (Monark 874E, Sweden). Saddle and handlebar heights/positions were adjusted to each subject’s riding comfort and replicated for each trial. A metabolic analyzer (Parvo Medics, Logan, UT) was used during all trials. Subjects wore a nose clip and expired through a two-way rebreathing valve connected to the metabolic cart. HR by telemetry was recorded simultaneously (Polar Electro Inc., Lake Success, NY), and all data were evaluated using 15 s averaging. Filter replacement and calibration were performed between tests according to manufacturer’s guidelines.
GXT with square wave verification bout
Demographic data, PA-R, and measurements of height and body mass were collected during the initial visit. A modified PA-R scale derived by George et al. (11) was used to accommodate a more robust population with respect to physical activity level (Table 1). Relative V˙O2max was estimated using a nonexercising regression equation (15) (Equation 2), and that value was subsequently used to estimate Wpeak as shown as follows (1):
where V˙O2max is expressed in milliliters per kilogram per minute, body mass (BM) is in kilograms, and Wpeak is in watts.
The estimated Wpeak for each subject was used to prescribe a custom GXT with a time limit estimate of 10 min using the following: Wpeak/10 min = 1 min stages (W·min−1) (25). A 3-min active recovery of 50 W was administered after the GXT. Subsequently, a square wave exhaustive bout performed two gradations below Wpeak was used to verify V˙O2max (25). Exhaustion for the GXT and square wave verification bout was defined as a decline in cadence by 10 rpm for 10 s despite verbal encouragement. The highest V˙O2 data collected during the GXT and square wave verification bout were compared to determine “true” V˙O2max values (within 1.5 mL·kg−1·min−1). Data from the metabolic analyzer were extracted, and detection of watts at gas exchange threshold (WGET) was determined using inflection of VCO2-V˙O2 (y-axis) relative to inflection of power output (x-axis) (25).
Mankato submaximal exercise test
Self-reported and demographic data were collected from the initial visit and were used to prescribe the MSET. Subjects performed the MSET at their preferred cadence determined during visit 1. The test consisted of two consecutive 3-min stages. In stage 1, all subjects cycled at 35% of estimated Wpeak (i.e., stage anticipated below GET), followed by second stage at either 65% or 70% of estimated Wpeak (i.e., stage anticipated above GET), which was dependent on subject’s PA-R. Subjects with a PA-R of 0 to 9 cycled at 65%, and those with a PA-R greater than or equal to 10 cycled at 70% (5,6).
YMCA cycle test
The YMCA test was conducted as described by The Y’s Way to Physical Fitness (12). Subjects performed two to four 3-min consecutive stages at 50 rpm. Subjects completed a fourth minute, if HRs from second and third minutes were not within 5 bpm. All subjects began their first stage at 25 W, the second stage was determined based on HR values from the first stage (<80 bpm: 125 W, 80–89 bpm: 100 W, 90–100 bpm: 75 W, >100 bpm: 50 W). Termination of the test occurred when 2 consecutive stages were completed with a HR between 110 and 150 bpm (12). Data collected from the last 2 stages of the YMCA test and stages of the MSET were used to extrapolate estimated Wpeak, via linear regression of power output (y-axis) and HR (x-axis). The APMHR was predicted using 208 − (0.7 × age) (32); V˙O2max was subsequently calculated using equation 1.
“True” V˙O2max was defined as the highest measured V˙O2 from either the GXT or exhaustive bout. A one-way analysis of variance with repeated measures was used to assess differences between the maximal GXT and each estimate of V˙O2max. Bonferroni adjustments were used to evaluate main effects. Consistency between measured and estimated values were evaluated using a two-way mixed intraclass correlation coefficient (ICC α), SEM (i.e., typical error), and coefficient of variation (CV) (14). Pearson product–moment correlation and standard error of estimate (SEE) values are reported to facilitate comparison of measurement agreement with previous research. Separate one-way analysis of variance with repeated measures was used to evaluate V˙O2 and HR values sampled every 30 s from the MSET. A Bonferroni adjustment was used to evaluate differences in V˙O2 and HR values between 30-s time intervals. A paired sample t test was used to compare WGET and prescribed power output during the second stage of the MSET. Descriptive statistics are reported using mean ± SD. Alpha for rejecting the null hypothesis was set at P ≤ 0.05.
V˙O2 values from the GXT and square wave verification bout were 44.4 ± 8.9 and 44.6 ± 8.9 mL·kg−1·min−1, respectively, (ICC = 0.99, SEM = 0.99 mL·kg−1·min−1, CV = 2.4%) with the highest value used to identify “true” V˙O2max (45.1 ± 8.7 mL·kg−1·min−1) (note: these values were similar if the average of the GXT and square wave bout was used as the criterion measure for V˙O2max.). There was a significant main effect for estimating V˙O2max, (F = 12.7, P < 0 .05). The MSET (43.6 ± 8.6 mL·kg−1·min−1) did not differ from “true” V˙O2max (P = 0.50), whereas the YMCA test (41.1 ± 9.6 mL·kg−1·min−1) yielded an underestimation (P = 0.002) (Tables 2 and 3). The MSET was correlated moderately with “true” V˙O2max (ICC = 0.72, CV of 11.3%). The YMCA test was correlated poorly with “true” V˙O2max (ICC = 0.27, CV of 15.2%) (Table 3). The nonexercising equation did not differ from “true” V˙O2max (46.8 ± 8.7 mL·kg−1·min−1 (P = 0.01); however, it was correlated moderately with “true” V˙O2max (ICC = 0.60, CV of 8.34%) (Table 3).
The r value and SEE for the YMCA test was r = 0.64, 6.91 mL·kg−1·min−1, whereas the MSET was r = 0.72, 6.20 mL·kg−1·min−1 (Fig. 1). The YMCA tests mean difference was −3.96 mL·kg−1·min−1, whereas the mean difference for the MSET was −1.54 mL·kg−1·min−1. The r value and SEE and for the nonexercising equation was 0.80 and 5.36 mL·kg−1·min−1, respectively. The power output selected for the second stage of the MSET was 17 W greater than GET (P < 0.05). V˙O2 data from stage 1 of the MSET depicts a steady state beginning at 1.5-min; V˙O2 data during minutes 1 and 1.5 did no differ (P < 0.05) (Fig. 2). V˙O2 data from stage 2 depicts a delayed steady state beginning at minute 5.5; V˙O2 data during minutes 5.5 and 6 did not differ (P < 0.05) (Fig. 2A); a similar trend was observed for HR (Fig. 2B).
To our knowledge, this is the first study to compare estimated V˙O2max from submaximal exercise testing to the “true” V˙O2max determined using a GXT and an exhaustive bout for verification. The principal findings of the present study are as follows. Our results demonstrate the MSET yields a more valid estimate of V˙O2max compared with the YMCA test. We believe these results are due to the fact the MSET uses demographic data and the nonlinear oxygen uptake-power output relationship.
Using demographic data and PA-R (Table 1), we estimated WGET through a series of three equations (equations 2–4.) (6,15). Indeed, the actual power output used in the MSET was approximately 17 W above WGET determined in the GXT. By selecting an intensity exceeding the WGET, we were able to discern a visible V˙O2 slow component (Fig. 2A).
There are a number of shortcomings associated with the YMCA test. The assumption of the YMCA test is that oxygen uptake-power output relationship is linear. Such an assumption is flawed because the relationship is nonlinear at intensities exceeding GET (4,16,19,24,33). The YMCA test equally assumes that the HR–power output relationship is linear during the YMCA test. Thus, for many people the YMCA test yields a low estimate of V˙O2max. In the present study, there was a time-dependent rise in HR relative to time (Fig. 2B), observed in the second stage of the MSET. Previous research has indicated changes in HR with changes in power output are greater for intensities above GET in comparison to intensities below GET (26).
The YMCA test prescribes predetermined intensities that fail to address differences in gender, PA-R, and age. Beekley et al. (2) observed gender differences for the YMCA protocol, including mean differences, were 9.5 and 3.9 mL·kg−1·min−1 for men and women, respectively (2). Observed mean differences from the YMCA test from the present study were 5.7 and 1.8 mL·kg−1·min−1 for men and women, respectively. Subjects from the present study exceeding 38 yr of age had mean difference of 2.8 mL·kg−1·min−1 for the YMCA test, whereas age did not influence the MSET (−0.2 mL·kg−1·min−1).
The nonexercising equation had a more favorable SEE, r value, typical error, and CV%, compared with either submaximal test and did not differ from “true V˙O2max” (Fig. 1 and Table 2). We must point out that the nonexercising consistently overpredicted V˙O2max, as demonstrated by the ICC (Table 3). Conversely, the MSET had a higher ICC than the nonexercising equation (0.73 versus 0.63, receptively).
To our knowledge, the present study is the first to test the validity of the YMCA test and MSET against a GXT with a square wave exhaustive verification bout to determine “true” V˙O2max. Widely used criteria for determining a successful GXT are as follows: V˙O2 plateau of 150 mL·min−1 or lower, an RER of 1.10 or greater, and APMHR of ±10 bpm (22). Previous research has refuted the need for a V˙O2 plateau to achieve a V˙O2max. For example, Day et al. (7) found that only 17% of their subjects achieved a V˙O2 plateau (7). Mier et al. (23) observed similar results where only 7 of the 35 college athletes tested achieved a V˙O2 plateau (23). Achieving an RER of 1.10 or greater has been refuted by previous research as well (21). Subjects either fail to obtain an RER of 1.10 or greater during long GXT durations (>13 min) or reach an RER of 1.10 or greater before reaching V˙O2max (17,28). Lastly, using APMHR to determine a successful GXT has been criticized. Poole et al. (28) demonstrated that subjects within their study achieved within 10 bpm of APMHR at 76% of V˙O2max. The present study demonstrated a CV from the GXT and submaximal verification bout was 2.4%, which is consistent with previous research (17,31).
Previous study by Beekley et al. (2) raised questions about the validity of the YMCA test and, to our knowledge, is the only group to validate the YMCA test against a cycle ergometer GXT. Interestingly, their prediction error was considerably larger than our study and overpredicted the YMCA test, demonstrating a mean difference of 6.83 versus −3.96 mL·kg−1·min−1) (2). Beekley et al. (2) used a different APMHR equation (220 − age) for the extrapolation of their YMCA test data; the validity of this equation has been questioned (30).
Our results indicate that the MSET is valid for cycle ergometry. Therefore, we recommend research using this procedure for other modes of exercise. A noteworthy difference would be identification of a suitable prediction equation to replace equations 1 to 4 for different modes. For example, researching predicted V˙O2max with a nonexercising equation for upper body ergometry.
There are a number of practical conveniences distinguishing MSET from the YMCA test. First, our protocol provides the freedom for the individual to select their preferred cadence. Second, the protocol is two stages and completed in 6 min, thus making the test time efficient. Third, the decision making related to setting each stage is determined before initiating the test, thus eliminating error. We have provided supplemental digital spreadsheet (http://links.lww.com/MSS/A574) so that the MSET may be administered by a wide range of practitioners.
The primary limitations of the study centered on our demographic restrictions. Most institutional review boards for the protection of human subjects are unfamiliar with the use of a verification bout for the determination of V˙O2max. Essentially, each subject is being “stressed” to their maximum effort twice within a single visit. Such a procedure, while emerging as a preferred method for determining “true” V˙O2max, is not included in leading position stands and cardiopulmonary exercise testing guidelines. Omission of the verification stage for V˙O2max assessment is indirectly limiting procedural advancements (e.g., our review board prohibited us from testing subjects with a BMI >30 who had an added risk factor).
Assigning a stage above GET will account for the nonlinear oxygen uptake-power output relationship thereby enhancing its validity. Application of the GXT with verification has emerged as the most appropriate method for assessing submaximal test validity. In conclusion, the MSET has greater validity than the YMCA test for estimating V˙O2max because of using demographic data for test administration.
Disclosure of Funding: No funding was received.
Conflict of Interest: No conflict of interest is present.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. ACSM. ACSM’s Guidelines for Exercise Testing and Prescription
. Lippincott Williams & Wilkins; 2013. pp. 19–173.
2. Beekley MD, Brechue WF, Dehoyos DV, Garzarella L, Werber-Zion G, Pollock ML. Cross-validation of the YMCA submaximal cycle ergometer test to predict VO2max. Res Q Exerc Sport
. 2004; 75 (3): 337–42.
3. Blair SN. Physical inactivity: the biggest public health problem of the 21st century. Br J Sports Med
. 2009; 43 (1): 1–2.
4. Brittain C, Rossiter H, Kowalchuk J, Whipp B. Effect of prior metabolic rate on the kinetics of oxygen uptake during moderate-intensity exercise. Eur J Appl Physiol
. 2001; 86 (2): 125–34.
5. Davis JA, Frank M, Whipp BJ, Wasserman K. Anaerobic threshold alterations caused by endurance training in middle-aged men. J Appl Physiol Respir Environ Exerc Physiol
. 1979; 46 (6): 1039–46.
6. Davis JA, Vodak P, Wilmore JH, Vodak J, Kurtz P. Anaerobic threshold and maximal aerobic power for three modes of exercise. J Appl Physiol
. 1976; 41 (4): 544–50.
7. Day JR, Rossiter HB, Coats EM, Skasick A, Whipp BJ. The maximally attainable VO2 during exercise in humans: the peak vs. maximum issue. J Appl Physiol (1985)
. 2003; 95 (5): 1901–7.
8. DiMenna FJ, Jones AM. “Linear” versus “nonlinear” VO2
responses to exercise: reshaping traditional beliefs. J Exerc Sci Fit
. 2009; 7 (2): 67–84.
9. Farrell SW, Fitzgerald SJ, Mcauley PA, Barlow CE. Cardiorespiratory fitness, adiposity, and all-cause mortality in women. Med Sci Sports Exerc
. 2010; 42 (11): 2006–12.
10. Foss Ø, Hallén J. The most economical cadence increases with increasing workload. Eur J Appl Physiol
. 2004; 92 (4–5): 443–51.
11. George JD, Stone WJ, Burkett L. Non-exercise V˙O2max
estimation for physically active college students. Med Sci Sports Exerc
. 1997; 29 (3): 415–23.
12. Golding L, Sinning W. Y’s Way to Physical Fitness: The Complete Guide to Fitness Testing and Instruction: YMCA of the USA
. Champaign, IL: Human Kinetics Publishers; 1989. pp. 89–106.
13. Hill DW, Vingren JL. The effect of pedalling cadence on maximal accumulated oxygen deficit. Eur J Appl Physiol
. 2012; 112 (7): 2637–43.
14. Hopkins WG. Measures of reliability in sports medicine and science. Sports Med
. 2000; 30 (1): 1–15.
15. Jackson AS, Blair SN, Mahar MT, Wier LT, Ross RM, Stuteville JE. Prediction of functional aerobic capacity without exercise testing. Med Sci Sports Exerc
. 1990; 22 (6): 863–70.
16. Jones AM, Poole DC. Oxygen Uptake Kinetics in Sport, Exercise and Medicine
. Routledge; 2013.
17. Kirkeberg J, Dalleck L, Kamphoff C, Pettitt R. Validity of 3 protocols for verifying VO2 max
. Int J Sports Med
. 2011; 32 (4): 266–70.
18. Lee DC, Sui X, Artero EG, et al Long-term effects of changes in cardiorespiratory fitness and body mass index on all-cause and cardiovascular disease mortality in men: the Aerobics Center Longitudinal Study. Circulation
. 2011; 124 (23): 2483–90.
19. Londeree B, Moffitt-Gerstenberger J, Padfield JA, Lottmann D. Oxygen consumption of cycle ergometry is nonlinearly related to work rate and pedal rate. Med Sci Sports Exerc
. 1997; 29 (6): 775–80.
20. Lucia A, Juan A, Montilla M, et al In professional road cyclists, low pedaling cadences are less efficient. Med Sci Sports Exerc
. 2004; 36 (6): 1048–54.
21. Midgley AW, Carroll S, Marchant D, McNaughton LR, Siegler J. Evaluation of true maximal oxygen uptake based on a novel set of standardized criteria. Appl Physiol Nutr Metab
. 2009; 34 (2): 115–23.
22. Midgley AW, McNaughton LR, Polman R, Marchant D. Criteria for determination of maximal oxygen uptake: a brief critique and recommendations for future research. Sports Med
. 2007; 37 (12): 1019–28.
23. Mier CM, Alexander RP, Mageean AL. Achievement of V˙O2max
criteria during a continuous graded exercise test and a verification stage performed by college athletes. J Strength Cond Res
. 2012; 26 (10): 2648–54.
24. Özyener F, Rossiter H, Ward S, Whipp B. Influence of exercise intensity on the on- and off-transient kinetics of pulmonary oxygen uptake in humans. J Physiol
. 2001; 533 (3): 891–902.
25. Pettitt RW, Clark IE, Ebner SM, Sedgeman DT, Murray SR. Gas exchange threshold and VO2max testing for athletes: an update. J Strength Cond Res
. 2013; 27 (2): 549–55.
26. Pettitt RW, Symons JD, Taylor JE, Eisenman PA, White AT. Adjustment for gas exchange threshold enhances precision of heart rate-derived VO2
estimates during heavy exercise. Appl Physiol Nutr Metab
. 2008; 33 (1): 68–74.
27. Pierre S, Nicolas H, Frédérique H. Interactions between cadence and power output effects on mechanical efficiency during sub maximal cycling exercises. Eur J Appl Physiol
. 2006; 97 (1): 133–9.
28. Poole DC, Wilkerson DP, Jones AM. Validity of criteria for establishing maximal O2
uptake during ramp exercise tests. Eur J Appl Physiol
. 2008; 102 (4): 403–10.
29. Pringle JS, Doust JH, Carter H, Tolfrey K, Jones AM. Effect of pedal rate on primary and slow-component oxygen uptake responses during heavy-cycle exercise. J Appl Physiol (1985)
. 2003; 94 (4): 1501–7.
30. Robergs RA, Landwehr R. The surprising history of the “HRmax = 220-age” equation. J Exerc Physiol
. 2002; 5 (2): 1–10.
31. Sedgeman D, Dalleck L, Clark IE, Jamnick N, Pettitt R. Analysis of square-wave bouts to verify VO2max. Int J Sports Med
. 2013; 34 (12): 1058–62.
32. Tanaka H, Monahan KD, Seals DR. Age-predicted maximal heart rate revisited. J Am Coll Cardiol
. 2001; 37 (1): 153–6.
33. Wasserman K, Whipp BJ. Excercise physiology in health and disease. Am Rev Respir Dis
. 1975; 112 (2): 219–49.
34. Zoladz JA, Duda K, Majerczak J. Oxygen uptake does not increase linearly at high power outputs during incremental exercise test in humans. Eur J Appl Physiol Occup Physiol
. 1998; 77 (5): 445–51.