### Abstract

VO2max is an invaluable measure for the assessment of aerobic fitness; however, to yield accurate results direct assessment requires costly equipment, trained investigators, and that the participant produce a maximal effort to volitional fatigue. The majority of VO2max prediction equations have attempted to predict aerobic capacity without considering physiological variables other than age and body composition. As a result, a majority of VO2max prediction equations have been found to be invalid. A recent study proposed an equation accounting for additional physiological variables known to influence aerobic capacity, including blood volume, fat-free mass, urinary creatine excretion, and total body potassium. Therefore, this investigation sought to evaluate the validity of novel non-exercise prediction equations, which utilize bioelectrical impedance analysis (BIA) to obtain an estimate of blood volume and skeletal muscle mass as predictor variables in an attempt to increase the accuracy of non-exercise VO2max prediction equations. VO2max was assessed using indirect calorimetry. Healthy male (30.9 ± 6.0 y, 179.0 ± 4.3 cm, 94.1 ± 19.5 kg; n = 23) and female (32.0 ± 6.1 y, 167.8 ± 7.9 cm, 72.0 ± 9.6 kg; n = 25) participants completed a VO2max test and a physical activity survey (PA-R) and were analyzed using bioelectrical impedance. Results indicated that each equation resulted in a significant (p ≤ 0.025) underestimation of VO2max. These outcomes suggest that the use of BIA to estimate blood volume and skeletal muscle mass does not improve the accuracy of VO2max prediction equations. Coaches and trainers will not benefit from the inclusion of BIA in an equation to predict aerobic fitness. Currently, the best methods to estimate aerobic fitness require submaximal and maximal exercise testing. Predicting aerobic fitness using non-exercise equations does not appear to be practical or valid.

**Keywords:** maximal, aerobic capacity, prediction, gender-specific

### Introduction

The rate of maximal oxygen consumption (VO2max) has practicality in research and field settings as a measure of aerobic fitness, in order to prescribe exercise intensities and to assess exercise training responses following an intervention (19). An acceptable standard for VO2max determination is the direct measure of expired gas samples obtained while an individual is performing maximal exertion exercise (2). From a research perspective reliable non-exercise VO2max prediction equations could prove to be beneficial, as experimenters could obtain an immediate, valid measure of the aerobic fitness of an individual without maximal exercise testing. Additional advantages of non-exercise VO2max prediction equations include the ease and cost associated with test administration and use in participants who are unable to perform a treadmill test, as VO2max tends to be underestimated with other modes of exercise (19). However, the greatest advantage of an accurate VO2max prediction equation is the practicality of use in research laboratories that do not possess the necessary equipment to access VO2max and for coaches and trainers looking to evaluate several athletes and/or an entire team. Due to the disadvantages associated with VO2max testing numerous submaximal (1,8,18,23) and non-exercise prediction equations (4,5,10,17,21,24,25) have been developed to reduce the necessity of direct VO2max assessment.

Previous non-exercise prediction equations have been developed but the need to improve the accuracy of these equations has been suggested in previous literature (4,16,17,21). However, due to known deviations in VO2max values determined from varying modes of exercise (bike, treadmill walking, treadmill running, and arm ergometry), the use of VO2max prediction equations are dependent on the task. For example, a prediction equation for VO2max during a treadmill run may not be accurate for predicting VO2max during cycle ergometry. In addition, another primary shortcoming of non-exercise VO2max prediction equations is the limited ability to account for genetic variability in VO2max (21). According to Stahn et al. (21), the primary physiological determinants measured at rest to predict VO2max are blood volume, which has been found to account for up to 80% of the variance in VO2max, and a group of variables including fat-free mass, urinary creatine excretion, and total body potassium, which have been proposed to be related to skeletal muscle mass. Additional evidence supporting this claim was provided by Sananda et al. (20) who found total skeletal muscle mass to be highly correlated (r = 0.92, p < 0.001) with VO2max (20).

Stahn et al. (21) sought to obtain an estimate of blood volume and skeletal muscle mass using bioelectrical impedance analysis (BIA). Previous work has suggested BIA to have a strong correlation with blood volume (r = 0.89, SEE = 9.0%) using the impedance index of height squared divided by impedance (22) and skeletal muscle mass, as compared to magnetic resonance imaging (r = 0.927, SEE = 9.0%) (11). As a result Stahn et al. (21) developed a non-exercise VO2max prediction equation, which utilizes BIA to estimate resting levels of blood volume and skeletal muscle mass as predictor variables. However, the equation by Stahn et al. (21) has yet to be validated by an independent laboratory, and the benefits of utilizing BIA for predicting VO2max have not been established. Therefore, the purpose of this study was to validate treadmill VO2max predictions using the recently published BIA equation of Stahn et al. (21). It was hypothesized that the BIA equations would produce accurate VO2max predictions due to the relationship between VO2max, BIA, skeletal muscle mass, and blood volume.

### Methods
#### Subjects

Sixty participants chose to participate in this study, but 12 were eliminated for not reaching VO2max (n = 48; Table 1). All testing was conducted after the participant signed the IRB-approved informed consent and completed comprehensive medical history questionnaires. Participants were excluded if they: 1) had a history of metabolic, hepatorenal, musculoskeletal, autoimmune, or neurological disease; 2) were currently taking androgenic medications; or 3) had consumed nutritional supplements that may affect metabolism [i.e., over 100 mg•d-1 of caffeine, ephedrine alkaloids, etc.] and/or muscle mass [i.e. creatine, protein/amino acids, androstenedione, dihydroepiandrosterone (DHEA), etc.] within three months of starting the study; 4) were unable to reach at least two of the three stated criteria for reaching VO2max.

Table 1. Participant characteristics of validated equations

Stahn et al. (21) Current Validation Participants
N Males Females N Males Females
N 66 33 33 48 23 25
Age (yr) 24.0 (4.0) 25.0 (4.0) 23.0 (4.0) 31.5 (6.0) 30.9 (6.1) 32.1 (6.1)
Height (cm) 174 (6) 180 (5) 168 (6) 173 (9) 179.0 (4) 169 (8)
Weight (kg) 68.4 (7.6) 74.9 (8.3) 61.8 (6.8) 82.6 (18.7) 94.1 (19.5) 72.0 (9.6)
PA-R 6.6 (1.1) 6.6 (0.9) 6.3 (1.3) 2.9 (1.9) 3.4 (2.3) 2.4 (1.5)
VO2max (ml*kg*min-1) 53.6 (5.0) 59.6 (5.5) 47.6 (4.4) 43.9 (13.4) 42.4 (14.4) 45.2 (12.6)

#### Non-Exercise VO2max Prediction Equations

The equations selected for validation were developed by Stahn et al. (21) and are presented in Table 2.

Table 2. Submaximal VO2max prediction equations

2MF Stahn et al. (21) VO2max (DF50) = 14.29 · H2/Z + 104.14 · PA-R – 440.79 • Gender (M = 1, F = 0) + 489.47
2M Stahn et al. (21) VO2max (DF50) = 14.29 • Height/Z + 104.14 • PA-R– 440.79 • Gender (M = 1) + 489.47
2F Stahn et al. (21) VO2max (DF50) = 14.29 • Height2/Z + 104.14 • PA-R – 440.79 • Gender (F = 0) + 489.47

∗ All values from prediction equations were converted to ml•kg•min-1
H = Height (cm)
Z = Impedance (Ohm)
PA-R = Physical activity rating scale
M = Male
F = Female

#### Experimental Design

Testing was performed between 9:00 a.m. and 3:00 p.m. in a temperature-controlled laboratory maintained at 21.6 ± 0.7oC and 28.2 ± 5.5% relative humidity. Prior to testing, each subject was instructed to avoid the consumption of alcohol, refrain from heavy exertion for 48 hours, and avoid smoking and caffeine consumption the day of testing. Subjects were also instructed to consume 2 liters of water the day before testing in an effort to promote normohydration.

#### Anthropometry and Physical Activity Assessment

After voiding their bladders, subjects changed into minimal clothing and removed footwear for measurement of body mass and height, conducted on a calibrated scale and stadiometer (Detecto, Webb City, MO). Body mass was measured to the nearest 0.2 kg and height was assessed to the nearest 0.5 cm. The PA-R was used to assess the average weekly physical activity patterns of each participant in the 6 months prior to testing (7).

### Bioelectric Impedance Measurement

Whole-body impedance measurements were performed using a single frequency (50 kHz) bioelectrical impedance analyzer (IMPTM DF50, ImpediMed Inc, Queensland, Australia). Each morning prior to testing, the bioelectrical impedance device was calibrated following the manufacturer’s guidelines. Measurements were taken from the right side of the body using a tetrapolar electrode arrangement following the standard procedures used by Stahn et al. (21). Prior to testing each subject was asked remove jewelry and excess clothing before being instructed to lie in a supine position for 10 minutes with arms and legs abducted from the body at 10˚ and 20˚ respectively, allowing body fluids to stabilize. Following identification of electrode placement, body hair was removed with a razor before the skin was cleaned with alcohol and allowed to dry. Current-inducing electrodes (575 mm2: 25 mm x 23 mm) (ImpediMed Electrodes, Queensland, Australia) were placed 1 cm below the phalangeal-metacarpal joint in the middle posterior surface of the hand and 1 cm below the transverse (metatarsal) arch on the dorsum of the foot. Detector electrodes of the same type were placed on the lateral epicondyle of the humerus and the lateral condyle of the femur according to the guidelines of Stahn et al. (21). Interclass and intraclass correlation coefficients for within and between days using this technology vary between 0.960 and 0.997 (6,21), while interindividual within-day reliability measures are commonly 1.3-2.0% (13,15,21).

#### VO2max Assessment

VO2max testing was performed on a calibrated Quinton treadmill (Q65 Series 90, Bothell, WA) according to Stahn et al. (21). Participants began the test with a 4-minute warm-up at 1.5 m·s-1 at a 1% gradient. Following warm-up, 3-minute testing periods began at speeds of 2.0 m·s-1 for women and 2.5 m·s-1 for men. Completion of each stage resulted in a speed increase of 0.5 m·s -1 until volitional fatigue despite verbal encouragement.

Maximal heart rate, respiratory exchange ratio (RER), and VO2max were measured with a calibrated metabolic cart (ParvoMedics TrueOne® 2400 metabolic measuring system, Sandy, UT). The system was calibrated 15 minutes prior to testing according to manufacturer specifications. Mean oxygen uptake (VO2), carbon dioxide output (VCO2), and pulmonary ventilation (VE) were computed for each breath and averaged over 15-second intervals. Heart rate was monitored during testing using a heart rate monitor (Polar F6, Lake Success, NY). The test was considered maximal if two of the following criteria were obtained: 1) a plateau of VO2 occurred, defined as an increase of less than 150 ml·min-1 despite increasing speed, 2) Respiratory exchange ratio (RER) was ≥ 1.15, and 3) maximal heart rate was within 10 beats of age-predicted maximal heart rate (21).

#### Data analysis

Validity of VO2max estimates were based on an evaluation of predicted values versus the criterion value from direct treadmill VO2max assessment by calculating the constant error (CE = actual VO2max – predicted VO2max), r value (Pearson product moment correlation coefficient), standard error of estimate and total error (9,14). The mean difference (CE) between the VO2max prediction equations and the direct measure of VO2max was analyzed using dependent t-tests with the Bonferroni alpha adjustment (12). The method of Bland and Altman (3) was used to identify the 95% limits of agreement between actual VO2max values and predicted VO2max values.

### Results

Demographic information of participants in the Stahn et al. (21) study and the current investigation are presented in Table 1. To optimize the accuracy of the prediction equations, results of the validation analysis are presented in two groups: male- and female-specific equations (Table 3). Each sex-specific equation produced a significantly different VO2max value from the direct measure (p<0.05). TE values were greater than 13.2 ml•kg•min-1, SEE values were greater than 9.1 ml•kg•min-1 and r values were less than 0.75.

Table 3. Validity of non-exercise prediction equations for estimating VO2max ml•kg•min-1

Method VO2max ± (x SD) CE r Slope Y-intercept SEE TE
Direct VO2M 42.4 (14.4)
Male 33.3 (8.3) 9.1* 0.74 1.2 -0.5 9.9 13.3
Direct VO2F 45.2 (12.6)
Female 34.0 (6.5) 11.2* 0.70 1.3 -0.78 9.2 14.5

### Discussion

The sex-specific equations analyzed in this investigation produced predicted VO2max values that were significantly below the actual VO2max (p<0.05). Using the predicted VO2max values to produce exercise prescriptions would yield exercise intensities underestimated by an equivalent amount.

The aim of the Stahn et al. (21) study was to demonstrate the viability of using BIA for the non-exercise prediction of VO2max. The authors attempted to account for the influence of physiological variables on aerobic performance by indirectly accounting for blood volume, fat-free mass, urinary creatine excretion and total body potassium with a time efficient assessment of blood volume and skeletal muscle mass using a BIA device. Results from the Stahn et al. (21) study appeared promising as their equation was reported to account for 88.7% of the variance in VO2max in an athletic population, and the authors postulated the equation would be more effective in a more diverse population. However, in the current investigation the equations developed by Stahn et al. (21) were found to be invalid in a population of healthy men and women. Errors in the equations were most likely introduced by using predicted values of blood volume and skeletal muscle mass (via BIA). In essence, predicted variables were used to predict another predictor, VO2max. The validity of the equations developed by Stahn et al. (21) may be improved by using a more accepted and still cost-effective measure of skeletal muscle mass, such as a multiple-site skinfold, as was used in VO2max prediction equations developed by Jackson et al. (10).

### Conclusions

The equation developed by Stahn et al. (21) may have been effective at predicting VO2max in the athletic population used in the original investigation but appears to significantly underestimate VO2max in a representative sample of healthy young men and women. Future prediction equations should include percent body fat and physical activity rating scales, as these variables appear to have the greatest predictive power in the estimation of non-exercise VO2max prediction equations. Although the prediction equations developed by Stahn et al. (21) were not found to be valid in this investigation, non-exercise VO2max prediction equations should attempt to increase their predictive power by accounting for physiological factors that are known to influence VO2max, namely skeletal muscle mass. Furthermore, future research should examine the accuracy of the equations developed by Stahn et al. (21) in an athletic population and determine the viability of using a BIA device in the prediction of VO2max.

### Applications in Sport

An athlete’s aerobic fitness is a crucial component of performance regardless of the sporting event. Aerobic athletes and coaches/trainers can benefit from accurate measurements of aerobic fitness through VO2max testing. However, direct VO2max testing requires expensive equipment and is not practical in the field. Many prediction equations have been developed in an attempt to find an easy way to predict VO2max in the field. However, results from this investigation suggest that using BIA in a non-exercise VO2max equation may not be appropriate or valid in healthy men and women. Specifically, the Stahn et al. (21) BIA VO2max equations underpredicted VO2max, resulting in significantly lower VO2max values, giving the impression of an individual who is less aerobically fit. Therefore, it is suggested that coaches and trainers utilize either submaximal or maximal VO2max prediction equations for their athletes and clients, as non-exercise prediction equations may not provide valid information.

### Acknowledgements

The authors would like to thank all the participants for their willingness to participate in this investigation.

### References

1. Astrand, & Ryhming, I. (1954). A nomogram for calculation of aerobic capacity (physical fitness) from pulse rate during sub-maximal work. J Appl Physiol, 7(2), 218-221.
2. Balady, G. J., Berra, K. A., Golding, L. A., Gordon, N. F., Mahler, D. A., Myers, J. N., et al. (2000). ACSM’s guidelines for exercise testing and prescription (6 ed.). Philadelphia: Lippincott, Williams & Wilkins.
3. Bland, J. M., & Altman, D. G. (1986). Statistical methods for assessing agreement between two methods of clinical measurement. Lancet, 1(8476), 307-310.
4. Davis, J. A., Storer, T. W., Caiozzo, V. J., & Pham, P. H. (2002). Lower reference limit for maximal oxygen uptake in men and women. Clin Physiol Funct Imaging, 22(5), 332-338.
5. Fairbarn, M. S., Blackie, S. P., McElvaney, N. G., Wiggs, B. R., Pare, P. D., & Pardy, R. L. (1994). Prediction of heart rate and oxygen uptake during incremental and maximal exercise in healthy adults. Chest, 105(5), 1365-1369.
6. Fornetti, W. C., Pivarnik, J. M., Foley, J. M., & Fiechtner, J. J. (1999). Reliability and validity of body composition measures in female athletes. Journal of Applied Physiology, 87(3), 1114-1122.
7. George, J. D., Stone, W. J., & Burkett, L. N. (1997). Non-exercise VO2max estimation for physically active college students. Med Sci Sports Exerc, 29(3), 415-423.
8. Golding, L. A. (2000). YMCA Physical Fitness Testing and Assessment Manual (4 ed.). Champaign: Human Kinetics
9. Heyward, V. H., & Wagner, D. R. (2004). Applied Body Composition Assessments. Champaign, IL: Human Kinetics.
10. Jackson, A. S., Blair, S. N., Mahar, M. T., Wier, L. T., Ross, R. M., & Stuteville, J. E. (1990). Prediction of functional aerobic capacity without exercise testing. Med Sci Sports Exerc, 22(6), 863-870.
11. Janssen, I., Heymsfield, S. B., Baumgartner, R. N., & Ross, R. (2000). Estimation of skeletal muscle mass by bioelectrical impedance analysis. J Appl Physiol, 89(2), 465-471.
12. Keppel, G. a. T. D. W. (2004). Design and Analysis: A Researchers Handbook (4th ed.). Upper Saddle River, NJ: Prentice Hall.
13. Kushner, R. F., & Schoeller, D. A. (1986). Estimation of total body water by bioelectrical impedance analysis. Am J Clin Nutr, 44(3), 417-424.
14. Lohman, T. G. (1996). Human Body Composition. Champaign, IL: Human Kinetics.
15. Lukaski, H. C., Johnson, P. E., Bolonchuk, W. W., & Lykken, G. I. (1985). Assessment of fat-free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr, 41(4), 810-817.
16. Malek, M. H., Berger, D. E., Housh, T. J., Coburn, J. W., & Beck, T. W. (2004). Validity of VO2max equations for aerobically trained males and females. Med Sci Sports Exerc, 36(8), 1427-1432.
17. Matthews, C. E., Heil, D. P., Freedson, P. S., & Pastides, H. (1999). Classification of cardiorespiratory fitness without exercise testing. Med Sci Sports Exerc, 31(3), 486-493.
18. Pare, G., Noreau, L., & Simard, C. (1993). Prediction of maximal aerobic power from a submaximal exercise test performed by paraplegics on a wheelchair ergometer. Paraplegia, 31(9), 584-592.
19. Ross, R. M. (2003). ATS/ACCP statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med, 167(10), 1451; author reply 1451.
20. Sanada, K., Kearns, C. F., Kojima, K., & Abe, T. (2005). Peak oxygen uptake during running and arm cranking normalized to total and regional skeletal muscle mass measured by magnetic resonance imaging. Eur J Appl Physiol, 93(5-6), 687-693.
21. Stahn, A., Terblanche, E., Grunert, S., & Strobel, G. (2006). Estimation of maximal oxygen uptake by bioelectrical impedance analysis. Eur J Appl Physiol, 96(3), 265-273.
22. Stahn, A., Terblanche, E., & Strobel, G. (2004). Relationships between bioelectrical impedance and blood volume. Proceedings of the 11th Pre-Olympic Congress, 219-220.
23. Storer, T. W., Davis, J. A., & Caiozzo, V. J. (1990). Accurate prediction of VO2max in cycle ergometry. Med Sci Sports Exerc, 22(5), 704-712.
24. Wier, L. T., Jackson, A. S., Ayers, G. W., & Arenare, B. (2006). Nonexercise models for estimating VO2max with waist girth, percent fat, or BMI. Med Sci Sports Exerc, 38(3), 555-561.
25. Williford, H. N., Scharff-Olson, M., Wang, N., Blessing, D. L., Smith, F. H., & Duey, W. J. (1996). Cross-validation of non-exercise predictions of VO2peak in women. Med Sci Sports Exerc, 28(7), 926-930.

### Corresponding Author

Jordan R. Moon, PhD
Department Head
Department of Sports Fitness and Health
United States Sports Academy
One Academy Drive
Daphne, AL 36526

### Author Affiliations

Jordan R. Moon, PhD
Department of Sports Fitness and Health
United States Sports Academy
One Academy Drive
Daphne, AL 36526

Chad M. Kerksick, PhD and Jeffrey R. Stout, PhD
Department of Health and Exercise Science
University of Oklahoma
1401 Asp Ave.
Norman, OK 73019

Vincent J. Dalbo, PhD
School of Medical and Applied Sciences
Institute of Health and Social Science Research
Central Queensland University
Rockhampton, Australia

Michael D. Roberts, PhD
University of Missouri-Columbia
Department of Biomedical Science
Veterinary Medicine Building
Columbia, MO 65211

Print Friendly, PDF & Email