Maximal workload prediction models in the clinical cardio-pulmonary effort test

Abstracts

OBJETIVO: Este estudo buscou derivar equações generalizadas para predição da carga máxima para homens e mulheres jovens. MÉTODOS: O método da ergoespirometria direta (Aerosport® TEEM 100, Estados Unidos da América do Norte) foi empregado para determinar o VO2máx e a carga máxima (Wmáx), no cicloergômetro (Monark®, Brasil), de 30 homens (25 ± 5 anos, 75,0 ± 10,7 kg; 48,4 ± 8,8 mL . kg -1 . min -1 e 243 ± 51 Watts) e 30 mulheres (26 ± 5 anos, 56,7 ± 5,9 kg, 39,8 ± 7,6 mL . kg -1 . min -1 e 172 ± 37 Watts). A idade e a massa corporal foram empregadas como variáveis independentes. Para todos os testes estatísticos aceitou-se o nível de significância de p < 0,05. RESULTADOS: No ajuste linear múltiplo a carga máxima foi explicada pela da idade e massa corporal em 54% (r = 0,73), para homens, e em 76% (r = 0,87), para mulheres, com erros padrões respectivamente de 0,66 W . kg -1 e de 25 Watts. As equações propostas passaram pela validação cruzada, empregando-se outra amostra com características similares de idade e VO2máx composta por 15 homens e 15 mulheres. A correlação intraclasse entre os valores de Wmáx preditos e os medidos através da ergoespirometria foram de 0,70 e 0,69, com erros padrões de 28,4 e 15,8 Watts, respectivamente, para homens e mulheres. CONCLUSÃO: Este estudo exibe equações generalizadas válidas para determinação da carga máxima no cicloergômetro para homens e mulheres.

Consumo de oxigênio; esforço físico; teste de esforço; limiar anaeróbio; limiar ventilatório

ORIGINAL ARTICLE

Maximal workload prediction models in the clinical cardio-pulmonary effort test

Fernando dos Santos Nogueira; Fernando Augusto Monteiro Sabóia Pompeu

Universidade Federal do Rio de Janeiro - UFRJ - Rio de Janeiro, RJ - Brazil

^{Mailing Address}

ABSTRACT

OBJECTIVE: This study sought to derive generalized equations for predicting maximal workload for young men and women.

METHODS: Direct ergospirometry (Aerosport^® TEEM 100, USA) was used to determine VO_2máx and the maximal work load (W_max) on the cycle ergometer test (Monark^®, Brazil) of thirty men (25 ± 5 years, 75.0 ± 10.7 kg; 48.4 ± 8.8 mL · kg ^{-1 .} min ^-1 and 243 ± 51 Watts) and thirty women (26 ± 5 years, 56.7 ± 5.9 kg, 39.8 ± 7.6 mL · kg ^{-1 .} min ^-1 and 172 ± 37 Watts). Age and body mass were used as independent variables. For all statistic tests, a p < 0.05 significance level was adopted.

RESULTS: In the multiple linear adjustment, the maximal workload was explained by age and body mass as 54% (r = 0.73) for men, and as 76% (r = 0.87) for women, with standard errors of 0.66 W · kg ^-1 and 25 Watts. The proposed equations were cross-validated using another sample with similar age and VO_2máx characteristics comprised of fifteen men and fifteen women. The intraclass correlation between the predicted W_max values and those measures by ergospirometry were 0.70 and 0.69, with standard errors of 28.4 and 15.8 Watts, respectively, for men and women.

CONCLUSIONS: This study exhibits valid generalized equations for determining the maximal cycle ergonometer workload for men and women.

Key words: Oxygen consumption, physicial exertion, stress test, anaerobic threshold, ventilatory threshold.

The term ergospirometry or spiroergometry was coined in 1929, suggested by Knipping & Brauer (see the comprehensive review by Hollmann & Prinz¹). The primary metabolic measurement in this test is the aerobic capacity (VO_2max). Maximal aerobic capacity (VO_2max) is the highest oxygen capture that an individual can attain during physical work when breathing air at sea level². This variable, according to the Fick principle, is determined by the cardiac output and the mixed arteriovenous difference of oxygen. The ergospirometric test, therefore, allows a valuable study of the integration between pulmonary, cardiovascular, and musculoskeletal systems^2,3, and in some cases is the only way to comprehend physiopathological mechanisms, such as in severe pulmonary vascular disease without direct hypertension, in the patent foramen ovale with left-right shunt development during exercise, in exertional dyspnea, and in exertional hypoxemia, among others⁴. Its application before invasive or high-cost procedures in large groups of patients with heart disease and lung disease has many advantages ^3,4.

The maximal effort cycle ergometric protocol with 1-min load increments was first proposed by Wasserman et al⁵, and later perfected by Buchfuhrer et al⁶. This protocol consists of continuous load increments added each minute in order to attain the maximal workload (W_max), characterized by objective criteria⁷ as 10 ± 2 min. Therefore, prediction of Wmax from variables available before the beginning of the study for 10% increments W_max per minute is necessary. The equations normally used^8-12 are not very accurate, since they come from populations with anthropometric, cardiopulmonary, and biomechanical characteristics different from those of Brazilians. Valuable experimental assays and normative reviews published by groups of Brazilian researchers, however, have failed to present an alternative to these equations^13-16. Because of this difficulty, this study aims to a) develop mathematical equations for predicting W_max in a continuous and escalating cycloergometer test with 10 ± 2 min duration, and b) test the external validity of the equations developed herein, confronting them with the validity of those derived for a foreign population.

METHODS

The subjects of this study were 90 apparently healthy, non-smoking and non-athletic adult volunteers, 45 men and 45 women, women (Table 1). Subjects were randomly divided into two groups similar in age and body mass; the number of subjects was established according to Hopkins¹⁷. Thirty men and thirty women were randomly selected for the internal validity group (IV), and fifteen men and fifteen women for the external validity group (EV). Subjects were advised to abstain from extenuating physical activities and to maintain a mixed diet on the day prior to the test (> 5 METs). It was also recommended that they avoid ingestion of food and caffeine for three hours before the exercise. Prior to the test, the subjects filled out an informed consent form. The procedures adopted were approved by the local Ethics Committee for studies with human subjects.

Test protocol - A continuous and maximal escalating protocol was used⁵, comprised of an initial resting period for six minutes with the patient sitting on the cycle ergometer (Monark^®, Brazil), followed by a four-minute warm-up of pedaling without any workload, and later, by the escalating phase with a maximum duration between eight and twelve minutes⁶. Any test not interrupted because of fatigue during this period of time was excluded. Overload increments were determined for a constant pace in an empirical and individualized manner. These increments were maintained throughout the test (approx. 25 W . min^-1). The pace was controled by means of an audiovisual metronome (approx. 74 rev . min^-1).

Gas exchange and ventilatory variables were measured by a metabolic analyzer (Aerosport^® TEEM 100, USA) in an open circuit and by a medium flow pneumotachometer (Hans Rudolph^®, USA). These data were recorded every twenty seconds. Heart rate was monitored throughout the test by a cardiotachometer (Polar^® Vantage NV, Finland).

Before each test, equipment was calibrated. The ergospirometer was calibrated in a closed circuit (AGA^®, Brazil), furnishing a gas mixture containing 17.01% oxygen, 5.00% carbon dioxide, and balanced with nitrogen. Gas flow was calibrated using a three-liter syringe (Hans Rudolph^®, USA) and the cycle ergometer by means of a 3 kg ballast.

Tests were considered maximal when at least three of the following criteria were observed⁷: plateau in VO₂ (increase < 150 mL . min^-1 or 2 mL . kg^-1 . min^-1), HR_max > 90% of predicted value for age (220 - age), Concept of Perceived Effort > 18, RER (respiratory exchange rate) > 1.15, and maximal voluntary fatigue with inability to maintain the preestablished rhythm. VO_2max was determined as the highest value found at the end of the test. In this study, the maximal workload was defined as that observed in VO_2max.

Statistical analysis - Statistical processing was done by means of the Statistical Package for the Social Sciences^® package (SPSS, USA) version 13.0 and Microsoft Excel^® for Windows XP^® (USA). Descriptive statistics were used with means ± standard deviations (SD). For Group IV, stepwise multiple regression analysis was used to deduce mathematical models for prediction of maximal workload expressed in Watts (W). Eight equations^8-12 (Chart 1 and 2) for the same purpose (male and female genders) were analyzed by cross-validation with the subjects from Group IV. Values predicted for Group IV by means of foreign equations, and the values measured were confronted by analysis of variance with one factor and a post-hoc Tukey-HSD test. External validity of the equations derived in this investigation were tested by applying them to the EV group. Bland-Altman¹⁹ and modified Bland-Altman²⁰ limits of agreement plots were used with the results measured and calculated by the predictive formulas. The degree of association between the values measured and those predicted was determined by the intraclass correlation coefficient (ICC)²¹. The error of prediction was also noted by means of the technical error of the measurement (s=S.D._dif÷ Ö2) and the coefficient of variation (CV %). All statistical tests were performed at the < 0.05 significance level.

RESULTS

Characteristics of VO_2max and W_max of the volunteers are displayed on Table 1. The equation for prediction of maximal workload derived for the male group was:

Equation 1

W · kg^-1=6.413–(0.0531 · age)–(0.0242 · weight)

R² = 0.54 and SEE = 0.66 W · kg^-1

For the female group:

Equation 2

W_máx=-115.756+(2.271 · age)+(4.043 · weight)

R² = 0.76 and SEE = 25.03 W

Where: R² = coefficient of determination and SEE = Standard Error of the Estimate

The prediction results using equations 1 and 2 derived in this study did not significantly differ from the results measured for the EV group (Fig. 1). A significant difference was detected for foreign equations between the values measured and those predicted, for both male and female groups (Fig. 1 and Chart 2).

Table 2 displays the results of the analysis performed with the foreign equations. Group IV had limits of agreement of 0.93 ± 95.0 W for the male group, and -20.5 ± 47.6 W for the female group (Fig. 2). These indexes for the EV group were 34.2 ± 86.0 W for the male group, and -4.14 ± 47.87 W for the female group (Fig. 3). There was a moderate association between the values measured and those predicted (ICC = 0.70 and 0.69), respectively, for men and women. The SEE for the male group was 28.42 W (CV = 10.45%), and for the female group, 15.78 W (CV = 9.55%).

DISCUSSION

We recommend a 10% progression of the maximal workload per minute in the escalating, continuous and maximal protocol on the cycle ergometer for the detection of the anaerobic threshold and VO_2max. The test should be finalized by maximal effort criteria⁶ after eight to twelve minutes⁶. Based on these requirements, different equations were presented for determining the W_max, normally using the anthropometric variables obtained at rest such as gender, age, weight, and stature. Despite the large number of prediction equations found in literature, none has proved to be well suited to the Brazilian population. It seems wise to expect a reasonable prediction capacity when the equation is applied to the population from which it was derived. The results of this study suggest a good external validity for the equations proposed herein when compared to equations coming from alien populations.

In rapid tests (< eight minutes), Buchfuhrer et al.,⁶ observed a low VO_2max value. This was possibly due to limitations of muscular force. However, a reduction of VO_2max in longer (> sixteen minutes) tests would also be expected because of an increase in core temperature, dehydration, discomfort, or fatigue of ventilatory muscles⁶. Since longer tests require more time, do not furnish additional information, and do not produce maximal values, protocols with 10 ± 2 minutes duration are recommended.

The precision of gas exchange and ventilatory measurements is essential for data to be reproduced, and quality control needs to be made by means of calibration, operation, and analysis performed by experienced technicians¹⁵. Tests in which these precautions are taken show a low variation in repeat measurements at the nearest timepoints^15,22-25. Daily intrasubject variation, due to error and physiological fluctuations of oxygen consumption, minute ventilation, and heart rate, are²⁵ 3.8%, 8.0%, and 3.0%, respectively. Granja Filho et al²², observed a 5.5% intrasubject variation for VO_2max in a study carried out recently. We conclude that this result is a fruit of our equipment for clinical use. Even with this greater inaccuracy, the ergospirometer adopted here was validated by another group²³ and is amply used in Brazilian laboratories because of its low cost. This difference in accuracy between measurements obtained with more sofisticated equipement and ours (3.8% versus 5.5%) may lead to a small error in the application of our formulas, which could not yet be determined.

Coefficients of determination for the prediction equations of this study were moderate for the male group (R² = 0.54) and high for the female group (R²= 0.76). We chose to express the maximal workload as W_max and W . kg^-1, respectively, for women and men. This adjustment in the dependent variable was necessary in order to obtain a greater predictive power in the second case. This is probably due to the lower body adiposity in the male group^26,27; therefore, a greater weight variation would imply a greater heterogeneity of lean body mass, and consequently, of muscular strength. These indexes are close to those of other studies^9-11. Ong et al¹¹, sought to improve the predictive model of W_max using logarithmic transformation. The use of a nonlinear model was proposed in order to correct the influence of age in the decline of physical conditioning. The research performed by Ong et al¹¹, however, did not obtain more precise results than other studies^8-10.

The SEE indexes for the equations in this study were similar to those of other equations^8-10. As a comparison, because of the difference in units recorded (W . kg^-1 ,W , kg . m . min^-1), the SEE was divided by the average of the group studied in order to express the SEE in a porcentage of the mean (SEE %). Jones et al,⁸ found 15% indexes for the male group and 26% indexes for the female group. On the other hand, Hsi et al,⁹ and Neder et al,¹⁰ found 14% and 17% values for the male group, and 12% and 13% values for the female group, respectively. In this study, these indices were 20% for the male group, and 14% for the female group. The indexes of SEE% for an indirect estimate of VO_2max in several field methods, such as the step test, submaximal cycle ergometer test, or walking/running on track test, represent 10% to 20% of the VO_2max measured²⁸.

No other author, among all those researched, performed a cross-validation study of his W^max prediction equations. We see on Table 2 that the intraclass correlation coefficients for external validity of the main equations analyzed in literature varied from ICC = 0.01 (6F) to ICC = 0.38 (2F). These indexes, as was expected, were lower than those found in the original assays and in the present study. Analyzing the proportion of possibly invalidated tests when these equations are used^8-12, there is a 40% to 64% loss of tests (Table 3). The equations we propose show 27% and 20% indexes for the male and female groups, respectively (Table 3).

Figures 4 and 5 present dispersion diagrams of the predicted values and the measured values for the equations with the lowest indexes of invalidated tests (1M and 1F). One can see that the inclination of the tendency line is near zero, suggesting an absence of covariance. Distortions such as these are common when the population validity of predictive models is considered^9,11.

Analysis of the limits of agreement reveals that there is a great variation between the equations. The total error (TE - the measured value minus the predicted value) varied considerably between the predictive models. A tendency to underestimate W_max when using the foreign models was noted. The present study, however, presented acceptable limits of agreement (Fig. 2). Some points showed considerable distancing as to the error in the male group. Nevertheless, for this analysis, one should consider the ± 2 min range previously recommended⁶ and the index of invalidated tests shown on Table 3. For the female group, a small overestimation of the measured value was noted.

The predictive technique error suggested is close to the variation range recommended by Buchfuhrer et al⁶. We conclude, therefore, that the inference equations of W_max derived in this study for the escalating and continuous cycle ergometer protocols, with eight to twelve minutes duration, can be used with satisfactory external validity. We also conclude that the equations derived for alien populations that we studied did not show fully satisfactory results for the local population. For the future, a repetition of this assay for other age brackets and for the diverse patient groups that could not be included here is recommended.

Acknowledgements

The authors of this study express their gratitude to the Associação dos Amigos do Centro de Estudos e Aperfeiçoamento do Hospital dos Servidores do Estado do Rio de Janeiro, in the person of Dr. Aluysio S. Aderaldo Jr., for the significant contribution toward this study and to the Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro - FAPERJ. We are also thankful for the valuable review by our colleagues, Verônica Salerno, Martha M. Sorenson, Ricardo Gancz, and Michelle Porto.

Potencial Conflict of Interest

No potential conflict of interest relevant to this article was reported.

REFERENCES

Received on 10/09/05

Accepted on 03/30/06

• 1. Hollmann W, Prinz JP. Ergospirometry and its history. Sports Méd. 1997; 23: 93-105.
• 2. Åstrand PO, Rodahl K, Dahl HA, Strømme SB. Evaluation of physical performance on the basis of tests. In: Textbook of Work Physiology. 4^th ed. Champaign: Human Kinetics; 2003: 273-97.
• 3. Jones NL. Clinical exercise testing. 4^th ed. Philadelphia: W.B. Saunders; 1997.
• 4. Wasserman K, Hansen JE, Sue DY, Whipp BJ, Casauri R. Principles of exercise testing and interpretation. 2^th ed. Philadelphia: Lea & Febiger; 1994: 2-8.
• 5. Wasserman K, Whipp BJ, Koyal SN, Beaver WL. Anaerobic threshold and respiratory gas exchange during exercise. J Appl Physiol. 1973; 35: 236-43.
• 6. Buchfuhrer MJ, Hansen JE, Robinson DY, Sue DY, Wasserman K, Whipp BJ. Optimizing the exercise protocol for cardiopulmonary assessment. J Appl Physiol. 1983; 55: 1558-64.
• 7. Howley ET, Basset Jr. DR, Welch HG. Criteria for maximal oxygen uptake: review and commentary. Med Sci Sports Exerc. 1995; 27: 1292-301.
• 8. Jones NL, Makrides L, Hitchcock C, Chypchar T, Mccartney N. Normal standards for an incremental progressive cycle ergometer test. Am Rev Respir Dis. 1985; 131: 700-8.
• 9. Hsi WL, Lan C, Lai JS. Normal standards for cardiopulmonary responses to exercise using a cycle ergometer test. J Formos Med Assoc. 1998; 97: 315-22.
• 10. Neder JA, Nery LE, Castelo A, Andreoni S, Lerario MC, Sachs A, et al. Prediction of metabolic and cardiopulmonary responses to maximum cycle ergometry: a randomized study. Eur Respir J. 1999; 14: 1304-13.
• 11. Ong KC, Loo CM, Ong YY, Chan SP, Earnest A, Saw SM. Predictive values for cardiopulmonary exercise testing in sedentary chinese adults. Respirology. 2002; 7: 225-31.
• 12. Wasserman K, Hansen JE, Sue DY, Whipp BJ. Protocols for exercise testing. In: Principles of exercise testing and interpretation. Philadelphia: Lea & Febiger; 1987: 58-71.
• 13. Guimarães GV, Bellotti G, Wajngarten M, et al. Exercício e insuficiência cardíaca. Estudo da relação da gravidade da doença com o limiar anaeróbio e o ponto de compensação respiratório. Arq Bras Cardiol. 1999; 73: 339-43.
• 14. Ribeiro JP. Limiares metabólicos e ventilatórios durante o exercício. Aspectos fisiológicos e metodológicos. Arq Bras Cardiol. 1995; 64: 171-81.
• 15. Guimarães JI, Stein R, Vilas-Boas F, et al. Normatização de técnicas e equipamentos para realização de exames em ergometria e ergoespirometria. Arq Bras Cardiol. 2003; 80: 458-64.
• 16. Yazbek Jr P, Carvalho RT, Sabag LMS, Batistella LR. Ergoespirometria: Teste de esforço cardiopulmonar, metodologia e interpretação. Arq Bras Cardiol. 1998; 71: 719-24.
• 17. Hopkins WG. A new view of statistics. Sportscience [online]. Available from: URL: http://sportsci.org/resourse/stats [Accessed 2005 Jan 5].
• 18. Malek MH, Berger DE, Housh TJ, Coburn JW, Beck TW. Validity of VO_2máx equations for aerobically trained males and females. Med Sci Sports Exerc. 2004; 36: 1427-32.
• 19. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986; 2: 307-10.
• 20. Hall C, Figueroa A, Fernhall B, Kanaley JA. Energy expenditure of walking and running: comparison with prediction equations. Med Sci Sports Exerc. 2004; 36: 2128-34.
• 21. Bartko JJ. The intraclass correlation coefficient as a measure of reliability. Psych Reports. 1966; 19: 3-11.
• 22. Granja Filho PCN, Pompeu FAMS, Ribeiro AP. A acurácia da determinação do VO_2máx e do limiar anaeróbio. Rev Bras Med Esporte. 2005; 11: 167-71.
• 23. Novitsky S, Degal KR, Chatr-Aryamontri B, Gubakov D, Katch VL. Validity of a new portable indirect calorimeter: the Aerosport TEEM 100. Eur J Appl Physiol. 1995, 70: 104/1-104/6.
• 24. Davies CTM, Tuxworth W, Young IM. Physiological effects of repeated exercise. Clin Sci. 1970; 39: 247-58.
• 25. Jones NL, Kane JW. Quality control of exercise test measurements. Med Sci Sports. 1979; 11: 368-72.
• 26. Pollock ML, Laughridge E, Coleman B, Linnerud AC, Jackson A. Prediction of body density in young and middle-aged women. J Appl Physiol. 1975; 38: 745-9.
• 27. Pollock ML, Hickman T, Kendrick Z, Jackson A, Linnerud AC, Dawson G. Prediction of body density in young and middle-aged men. J Appl Physiol. 1976; 40: 300-4.
• 28. Zwiren LD, Freedson PS, Ward A, Wilke S, Rippe JM. Estimation of VO_2máx: a comparative analysis of five exercise tests. Res Q Exerc Sports. 1991; 62: 73-8.

Publication Dates

• Publication in this collection
  28 Sept 2006
• Date of issue
  Aug 2006

History

• Received
  09 Oct 2005
• Accepted
  30 Mar 2006