Validity Of Linear Extrapolation, Cooper, And MST Methods In The Assessment Of Maximum Oxygen Uptake

Methods

Maximum oxygen uptake also referred to as Aerobic power (VO₂) is used to reference the maximum rate of oxygen consumption. Aerobic power is normally used as measure of aerobic or cardio-respiratory fitness. Due to the expensive nature of VO₂ computation with regard to cost and time spent on precise gas assessment; several predictive tests have been developed to aid in analysis of aerobic fitness. There are several tests that include performance-centric measures such as walking or running for a prescribed time frame, conducting a multistage progressive test (MST) which is characterised by speed increment every 2 minutes. During all these activities measurements of heart rate are take and extrapolated in order to come up with maximum heart rate that is then used to tabulate VO_2max (Nguyen, et al., 2013). The MST and Cooper test are considered to be the most effective when tackling a large group of participants; since, the two are maximal in nature they are flawed because they create potential health problems.

Moreover, both test demand complete dedication and motivation on the participants' part for them to provided truly maximal effort in the different activities/workload.  On the other hand, a submaximal test on a cycle ergometer does not require participants to give their all, but it is time consuming. Numerous studies have assessed the validity of the Copper, linear extrapolation, and MST separate but very few studies have compared the three tests within the same population. The major limitation observed in previous studies is the fact that correlation was only drawn between the direct measurement and the predicted score. Correlations are effective assessments of validity but they do not paint a comprehensive picture of the situation. The social aim of this study is to expound the existing knowledge on the validity of field testing. As such, the participants will provide the data that will be analysed to provide information on the validity of linear extrapolation, Cooper, and MST methods in the assessment of maximum oxygen uptake (Gross, et al., 2017).  

Thirty healthy university students were enrolled to take part in the data collection exercise on a volunteer basis. All participants willingly consented to their involvement in the study and the study was subsequently approved by the university ethics committee. The participants was exposed to several physical and sports activities; majority of which were endurance related with different levels of difficulty and training status to cater to the individual skills of the subjects.  

Results

Regardless of the year in which the assessment were performed (2014, 2015, 2016, or 2017), the subjects were expected to undertake four different tests across five subsequent days. On the first day the participants were expect to perform a submaximal cycle ergometer test followed shortly after (25-30 minutes) by a direct measurement on a cycle ergometer. These tests were conducted first because they could be conducted in the laboratory and it was relatively easily for the subjects' ECG results to be monitored before they were directed to more demanding sport activities (maximal tests). The Cooper test and MST were conducted on separate days in a non-chronological manner (to sustain randomness). The two tests were conducted with an interval of two days. The participants were expected to register heart rate of 120-170 beats/min after a three minute warm-up. The tests considered of four submaximal exercises conducted subsequently with each lasting five minutes; all four tests were performed on a Monark cycle ergometer. The cycling was conducted with regard to 60 revolutions per minutes and the participant's heart rate was measured during the fourth and fifth minutes of the exercise using a "ECG S&W Medical Cardio Aid" (Petrovic, 2016).

The expired air was collected from the participants shortly after their last minute of exercise. This collection was done with the aid of a Douglas bag that was attached to a removable rubber mouthpiece, standard tubing, and Has Rudolph 2700 Valve. The gases expired by the subjects were examined through the use of a Taylor Servomex 0₂ analyzer and the PK Morgan CO₂ analyzer. The volume of the expired gas was assessed using a Parkinson Cowan Volume meter. All the analyzers used in the testing phase were calibrated with regard to know gas concentrations prior to the performance of the various analytical tests. VO_2max was predicted with the aid of linear extrapolation to allow for the comparison of all three assessment techniques. Linear extrapolation was conducted as follows: the measured heart rate and submaximal VO₂ were both extrapolated to the maximum heart rate; the measured heart rate and submaximal VO₂ were both extrapolated to the predicted maximum heart rate;    the measured heart rate and submaximal VO₂ were both extrapolated to the predicted maximum heart rate (Kumar, et al., 2012).    

The participants were allowed a 25 to 30 minutes break between the end of the submaximal test and the start of the maximal cycle ergometer test. During this rest period the participants are familiarized with the cycle ergometer; mouthpieces and nose clips were also presented to the subjects. The maximal cycle ergometer testing protocols used in the study were based on recommendations given by the British Association of Sports Science. Therefore, as per the specifications of the submaximal cycle ergometer test, heart rate and oxygen consumption were assessed for each participant (Vickers, 2011).

Discussion

Three parameters were employed in the assessment of VO₂ and VO_2max. The first parameter is anthropometric measurement this involves the assessment of body mass using an equipment of the same name that has an accuracy metric to the 100g and can support up to 150 kilograms. While the participants height was assessed using a stadiometer with scaling intervals of 0.1 cm; these two variables (body mass and height) were used to compute the participants’ body mass index (Nguyen, et al., 2013). The second parameter is the “cardiopulmonary exercise testing protocol” which demands the use of MGC spirometer. This instrument measures the participants Oxygen Consumption (VO₂), the production of carbon (iv) oxide (VCO₂), pulmonary ventilation (VE), Oxygen ventilation equivalents (VE/VO₂), Carbon (iv) Oxide (VE/VCO₂), and finally respiratory exchange ratio (VCO₂/VO₂). The last parameter gas analysis is more of a follow-up to the second, it involves the assessment of the concentration of oxygen and carbon (iv) oxide being inhaled and exhaled by the participants (Nguyen, et al., 2013).

This test was also conducted in the same indoor sports arena, across a 234 meter long athletics track.  The warm-p exercise consisted of a combination of flexibility and aerobic activities. The exercise required participants to run in pairs which results in 15 sets of participants. The predicted VO_2max associated with distance run by the participants within the 12min duration was assessed with regard to the Cooper table. The individuals who managed to run more than 2 miles within the stipulated 12 minutes had their respective VO₂ predicted by introducing distance ran as an attribute in the Cooper regression equation (Adsiz, et al., 2016).  

Standard computation formulas were employed in the calculation of VO₂, VCO₂, tidal volume (V_T), HR, respiratory frequency (f), and minute ventilation (V_E) every 30 second after every minute of exercise. The values of V_E and V_T were represented with regard to BTPS, while VO and VCO₂ were illustrated with adherence to STPD. The formula used in the computation of predicted maximum/peak VO₂ was “Peak VO2=0.83*(1-0.007age)*(1-0.25S)”.  Where h_t is the height of participants in meters, S is a dummy variable for gender (S=0 for males and S=1 for females). At the end of each exercise the participants were asked to rate their level of breathing difficulty on the Borg Scale by pointing to the appropriate score value. The participants had been familiarized with the Borg scale prior to the commencement of the exercise activities. The Borg scale is used to assessment dyspnoea for each of the participant.  

Conclusion

Predicted peak HR=210-0.65age (years)

The statistical analysis was focused on the performance of regression assessment on all three tests namely (Cooper, MST, and Predicted L/E). The responsive variable was taken to be the predictive test while the cycle ergometer value was employed as the explanatory variable. For each of the three regressions the standard f-test for the slope of the line of best fit was equate to 1 and that of intercept was equaled to 0 (this was done to do away will biasness). In order to discern the effect of calibration, a measurement value of 60 ml.kg^-1.min^-1 was assigned to each test for the purposed of cycle ergometer value calculation. A 95% confidence interval assessment was performed through the use following simple formula (where A=intercept, B=slope, and S=Standard deviation of the regression) (Poole & Jones, 2012).     

The Study reveals that the mean (standard deviation) of the participants' age and body mass were 23 year (2.4 years) and 71 kilograms (4.6 kilograms). Participants' age and body mass intervals were 21-29 years and 64.5-94 kilograms.  Table 1 indicated below contains the mean (S.D.) for the various test in terms of ml.kg^-1.min^-1. The recorded maximum heart rate was employed in the linear extrapolation method. The scatter-plot Charts labelled 1 to 3 provide a linear relationship between the direct measurement of VO₂ on the cycle ergometer and values obtained as a result of the MST, Cooper, and predicted L/E tests (Vaquera, et al., 2015).  The charts reveal that there is an underestimation of VO₂ for predicted L/E and MST values. Since, the result was not significantly different from 1for predicted value of cycle ergometer VO₂ associated with the slope of the regression; then we can conclude that the bias for all predicted values can be assumed to be constant for measured cycle ergometer VO₂. Accordingly, estimates of correlation coefficients and systematic biases for each test are present in Table 2. The bias is tabulated by subtracting the cycle ergometer VO₂ from the respective test VO₂. And in the case of MST and predicted L/E there was a significant underestimation of systematic biasness (Vaquera, et al., 2015).  

Moreover, for all tests the predicted values correlated with the cycle ergometer VO_2max and the highest correlation (positive) was observed with the Cooper test. A very strong positive correlation and no systematic bias needs to be observed if an acceptable VO₂ is to derived, As such the correlation is required to be greater than 0.8. Moreover, the results in Table 2 prove that Cooper test is not only offers the best correlation but is also totally unbiased.  For all three tests 60 ml.kg^-1.min^-1 was assumed for both the corresponding cycle ergometer  value and the underlining test (e,g, MST) for each regression exercise (results in Table 3) (Vaquera, et al., 2015). According to results indicated in the table, both predicted L/E and MST have VO_2max for the cycle ergometer  values that is shifted significantly to right of the target value (60 ml.kg^-1.min^-1) and likewise they have expansive prediction range. On the other hand, the Cooper test to a fair degree satisfied the target value. Across the various measurements point MST and predicted L/E demonstrated considerable biasness and sizable variability. But the Cooper test was not adversely affected by these issues. Table 4 provides a comparison between predicted and measured submaximal VO₂ associated with the cycle ergometer. The table clearly indicates the percentage differences between predicted and measured submaximal VO₂. Nevertheless, the difference was found not to be statistically significant. (Vaquera, et al., 2015) 

Limitations

Tables and Figures

Cycle ergometer  (VO2max)	Cooper (VO2max)	MST (VO2max)	Predicted L/E (VO2max)
60.1 (8.0) ml.kg-1.min-1	60.6 (10.3) ml.kg-1.min-1	56.6 (8.0) ml.kg-1.min-1	52.0 (8.4) ml.kg-1.min-1

Table 1: means (standard deviation) in terms of ml.kg^-1.min^-1

Test	Correlation with cycle ergometer  VO2max	Bias (s.e.) to VO2max	Significance of Bias (p-value)
Cooper	0.92	+0.5 (0.9)	0.6
MST	0.86	-4.5 (0.9)	<0.0001
Predicted L/E (a)	0.76	-7.8 (1.4)	<0.0001
Predicted L/E (b)	0.74	-8.2 (1.3)	<0.0001
Predicted L/E (c)	0.76	-8.4 (1.6)	<0.0001

 Table 2: Correlation and Biasness results

Cooper	(53,66)
MST	(56,75)
Predicted L/E	(56,86)

Table 3: Predictive performance at target Value of tests 60 ml.kg^-1.min^-1

Workload	Percent Difference means (s.d.)
1	5.2 (18.0)
2	-0.3 (11.1)
3	1.9 (5.7)

Table 4: Comparison between Measure and predicted maximal VO₂ 

Chart 1: relationship between Cooper test and Cycle ergometer VO_2max 

Chart 2: relationship between MST test and Cycle ergometer  VO_2max 

Chart 3: relationship between Predicted L/E test and Cycle ergometer VO_2max 

The direct measurement of cycle ergometer VO_2max which was 60.1 (8.0) ml.kg^-1.min^-1 is a clear representation of a fit group of individuals. This result is similar to the one recorded by McCutcheon in a 1990s that found 61.5 (7.1) ml.kg^-1.min^-1 in a cohort of sport players. It is however, important to note that the group is not homogeneous; this is due to the fact that VO_2max ranges between 46.9 and 76.3 ml.kg^-1.min^-1 (Abdelkrim, et al., 2010).

This study demonstrated that the Cooper test has a correlation coefficient of 0.92 with cycle ergometer VO_2max. As such, the results of this study are considerably similar to those of Cooper and McCutcheon which are 0.90 and 0.84 respectively. Other research studies over the years have shown this correlation coefficient to range between 0.7 and 0.9 (Ardigò, et al., 2018). Majority of this studies have used research participants of different gender, education level, and age set; for instance, 14 and 15 year adolescents, female athletes, and college graduates. The results of the study are influenced by age variation, status, number of subjects, and homogeneity or heterogeneity of participants. In the original research assessment conducted by Cooper, he evaluated 115 U.S. Air Force personnel with mean age of 22 and interval range of 17 to 54 years. The resultant VO_2max for all participants ranged between 31 and 59 ml.kg^-1.min^-1. 54% of all participants in this study were able to complete 2 miles in less than 12 minutes will none of subjects in cooper's test were able to go beyond this distance in the specified timeframe. Statisticians believe that in a group with considerable variation in age will result in a significant variation between VO_2max. This considerable difference between the highest and lowest VO_2max will contribute to be significantly high correlation coefficient (Ardigò, et al., 2018).     

In his research study, McArdle noted several problems that could be associated with comparing direct measurement and predictive tests with VO_2max in a homogenous group. For instance, he noted if the original Cooper test was limited to college student the correlation who be reduced from 0.59 from 0.90 (Adsiz, et al., 2016). Even thou the correlation results in this study and the original Cooper research are similar there are several differences between the overall conclusion. For instance, a hypothesis test revealed that unlike in the original Cooper research assessment, this study lacks systematic bias (with regard to the Cooper test). For instance, McCutcheon revealed that the copper test does sometimes attribute towards the systematic underestimated of 4 ml.kg^-1.min^-1 with regard to both female and male participants (were the mean age of the participants is 25 years). For the more than 50% of participants who were able to run more than 2 miles within 12 minutes their data was subject to the Cooper regression equation (Ardigò, et al., 2018).   

References

Abdelkrim, N. B. et al., 2010. Activity Profile and Physiological Requirements of Junior Elite Basketball Players in Relation to Aerobic-Anaerobic Fitness. The Journal of Strength and Conditioning , XXIV(9), pp. 2330-2342.

Adsiz, E., Nalcakan, G., Varol, S. R. & Vural, F., 2016. Determination of maximal oxygen uptake through a new basketball-specific field test. British Journal of Sports Medicine , XXIV(1), pp. 1-24.

Ardigò, L. P. et al., 2018. Effect of Heart rate on Basketball Three-Point Shot Accuracy. Frontiers in Physiology, pp. 1-54.

Gross, I. et al., 2017. Physical activity and maximal oxygen uptake in adults with Prader-Willi syndrome.. NCBI, pp. 1-54.

Kumar, S. K., Khare, P., Jaryal, A. K. & Talwar, A., 2012. Validity of Heart Rate Based Nomogram fors Estimation of Maximum Oxygen Uptake in Indian Population. All India Institute of Medical Sciences,, LVI(3), p. 279–283.

Nguyen, A. et al., 2013. Comparison of three methods for predicting the maximal oxygen uptake in a population of submariners. Elsevier Masson, LVI(1), pp. 315-321.

Petrovic, M., 2016. Biomechanics and the metabolic cost of walking in people with diabetes, Leuven: Katholieke Universiteit Leuven.

Poole, D. & Jones, A. M., 2012. Oxygen Uptake Kinetics. Comprehensive Physiology, II(2), pp. 933-996.

Vaquera, A. et al., 2015. Validity and Test-Retest Reliability of the TIVRE-Basket Test for the Determination of Aerobic Power in Elite Male Basketball Players. Journal of Strength Conditioning Research, XXX(2), pp. 12-35.

Vickers, R. R., 2011. Measurement Error in Maximal Oxygen Uptake Test. Naval Health Research Clinic, pp. 1-29.