Comments

Marais et al. (1999) investigated the response of selected physiological parameters to a variety of power outputs during upper-body arm crank ergometry. We write with interest regarding the calculation of energy expenditure and efficiency data presented in the article.

The first concern relates to the caloric equivalent of oxygen consumption at a respiratory exchange ratio (R) of 1.0 used in the study. In the article Marais and colleagues point out that most values of R recorded were in excess of 1.0. In view of this the authors used '... a mean caloric equivalent of 4.825 kcal × L^-1 oxygen consumed ...'. However, it is well documented that the caloric equivalence of oxygen consumption at and above an R-value of 1.0 ranges from 5.047 kcals × L^-1 (McArdle et al. 1996, [1]) to 5.189 kcal × L^-1 oxygen consumed (Pérronet & Massicotte 1991, [2]).

This fact leads to our second point. With the above information considered the values of energy expenditure (EE, presented in watts [W]) are rendered inaccurate. Indeed it would appear that the principles used by the authors to calculate EE are unclear. Using the mean value of V˙O₂ (2.0 L × min^-1) recorded at 50 % MP (118.4 W), Marais and colleagues reported a mean EE value of ∼ 175 W (Fig. 2, lower panel). Given that 1 W equates to 1 joule per second (1 J × s ^-1), the actual value of EE should have been in the region of 723 W. This value of EE has been calculated for a V˙O₂ of 2.0 L × min^-1, assuming an R of 1.0 and using the caloric equivalent of 5.189 kcal × L^-1 oxygen consumed

i.e. 2.0 L × min^-1 × 5.189 kcals × L^-1× 4182 J ÷ 60 s = 723.3 J × s^-1.

Even using the caloric equivalent for oxygen of 4.825 kcals × L^-1 oxygen consumed, the estimated value (673 W) remains considerably higher than that of the reported value (∼ 175 W). Please could the authors therefore clarify the method used in their calculations.

It is difficult to establish precisely why the inaccuracies associated with the calculation of EE have been made. However, the reported mean values of EE included in Fig. [2] (lower panel) appear to be in the order of one quarter of what they should actually be based on reported mean values of V˙O₂ across the full range of power outputs. It is therefore possible that the authors have not converted the caloric equivalent of V˙O₂ from kcals × L^-1 × min^-1 to kJ × L^-1 × min^-1 (i.e. to multiply kcals by a factor of 4.182) prior to making their final estimations of EE.

One further point of concern relates to the calculation and reported values of gross efficiency (GE, %). These data are presented in the upper panel of Fig. 2. Here, GE ranged from ∼ 16 % to 20 % across the full range of exercise intensities. However, if the mean values of power output and EE at 50 % MP are once again used to exemplify our point (i.e. 118.4 W and ∼ 175 W, respectively), a GE of ∼ 67 % is calculated - lying considerably outside the normal physiological range. However, the authors reported that the mean value of GE corresponding to the 50 % MP was ∼ 19.5 % - this information has been extracted from the upper panel of Fig. [2]. Once again a plausible reason for this discrepancy is not immediately apparent and requires clarification.

Finally, based on mean values of V˙O₂ presented by Marais et al., we have generated our own data table that displays the extent of differences reported for EE (ΔEE) and GE (ΔGE). This has been done for the self-selected (T_SUB.S) crank rate and for those both 10 % below and above (T_SUB-10% and T_SUB+10%, respectively) T_SUB.S. Furthermore, data has been calculated and presented for all crank rates across all relative power outputs used in the study (i.e. 50 %, 60 %, 70 %, and 80 % MP, respectively).

It is accepted that the issues highlighted above are 'relative' by nature and therefore do not influence the conclusions drawn by the authors. However, we would welcome a response from the workers clarifying precisely how they came to their final reported calculations with particular regard to estimations of EE and subsequent values of GE.

References

• 1 McArdle W D, Katch F I, Katch V L. Exercise Physiology: Energy, Nutrition and Human Performance. Baltimore; Williams and Wilkins 1996: 147
• 2 Pérronet F, Massicotte D. Table of Nonprotein Respiratory Quotient: An Update.  Can J Sp Sci. 1991;  16 23-29