Inside OUES: fact or fiction?

The ‘efficiency of the peripheral oxygen uptake’ or ‘oxygen uptake efficiency slope’, the so-called OUES, is defined by the logarithmic regression curve between oxygen uptake (VO2) and ventilation (VE). In other words, it is the ‘a’ in the VO21⁄4 a logVEþ b equation. In Figure 1, the VO2 versus VE relationship of a normal subject and of a patient with pulmonary hypertension is reported in the upper panel, and the OUES of the same subjects is shown in the lower panel. Steeper slopes represent a more efficient oxygen uptake. However, it seems to us that the physiological meaning of ‘efficiency of the peripheral oxygen uptake’ is undefined. This term needs at least some clarification and some physiology background. Indeed, efficiency of VE is how well or poorly VE performs. The main duty of VE is gas exchange for both oxygen and carbon dioxide. Therefore, VE efficiency is the highest when VE is the lowest for a given carbon dioxide uptake (VCO2) or VO2, provided that arterial carbon dioxide tension (PaCO2) or arterial oxygen tension (PaO2) are still. VE efficiency can be assessed for carbon dioxide as the slope of the relationship between VE and VCO2 measured from the beginning of loaded exercise up to the respiratory compensation point, when the slope is linear. Alternatively, VE efficiency can be assessed as the ratio of VE/VCO2 during exercise, usually at the anaerobic threshold or when it is at its lowest. And what about oxygen? Notably, the VE/VCO2 and VE/VO2 ratios are usually plotted in combination, and their changes are similar with a similar nadir, although the VE/VO2 ratio increases at the anaerobic threshold, while the VE/VCO2 ratio increases at the respiratory compensation point. Indeed, as opposed to a linear increase of VO2 during a progressively increasing workload exercise, the VE versus VO2 relationship is linear only up to the anaerobic threshold, i.e. for a small part of the exercise. Therefore, the slope of the VE versus VO2 relationship can be calculated as a linear equation only at the beginning of exercise, when VE is more erratic and more affected by psychological stress. In the late 1990s, Baba et al. suggested that the equation be linearised by applying the logarithmic regression, but this implies that the curve is exponential. Is that true? Probably not, because VE in an exercise ramp protocol from rest to peak is progressively under different domains – VO2, VCO2, PH and temperature, respectively. Notably, all of these domains are characterised by specific VE kinetics. Indeed, we identify thresholds on VE, such as the anaerobic threshold and the respiratory compensation point, both undefinable in an exponential equation. However, the true question is: does OUES tell something different from VE versus VCO2? Indeed, VO2 includes three major independent diffusion steps limiting oxygen transport: lungs, blood, and capillary to mitochondria. But almost the same applies to carbon dioxide, albeit in a reverse order: muscles to capillary, blood and lungs. However, the pressure gradients of oxygen and carbon dioxide between upper airways and mitochondria are different, as are their degrees of resistance to gas flow. Figure 2 shows that pO2 decreases from 160 mmHg in inspired air to 100 mmHg in the alveoli and almost to zero in the mitochondria, where it may even become negative in some circumstances, such as during heavy exercise. The pressure gradient for carbon dioxide is much lower, from approximately 45 mmHg in the muscles to 40 mmHg at the end of expiration. The reduction of pO2 has two quantitatively different steps, the smallest inside the lung and the largest from the capillary to the mitochondria. It is different for pCO2 changes, which are almost negligible, from the muscles, where carbon dioxide is stored, to venous blood, and from venous blood to expired air. As a matter of fact, end-tidal carbon dioxide is mainly an index of dead space VE, and therefore of VE efficiency. In conclusion, VE versus VCO2 is mainly lung efficiency, while VE versus VO2 is a composite of lung, circulation and capillary-to-mitochondria resistance to oxygen flow, i.e. total body efficiency, and not only peripheral

The 'efficiency of the peripheral oxygen uptake' or 'oxygen uptake efficiency slope', the so-called OUES, is defined by the logarithmic regression curve between oxygen uptake (VO 2 ) and ventilation (V E ). In other words, it is the 'a' in the VO 2 ¼ a Â logV E þ b equation. In Figure 1, the VO 2 versus V E relationship of a normal subject and of a patient with pulmonary hypertension is reported in the upper panel, and the OUES of the same subjects is shown in the lower panel. Steeper slopes represent a more efficient oxygen uptake. 1 However, it seems to us that the physiological meaning of 'efficiency of the peripheral oxygen uptake' is undefined. This term needs at least some clarification and some physiology background. Indeed, efficiency of V E is how well or poorly V E performs. The main duty of V E is gas exchange for both oxygen and carbon dioxide. Therefore, V E efficiency is the highest when V E is the lowest for a given carbon dioxide uptake (VCO 2 ) or VO 2 , provided that arterial carbon dioxide tension (PaCO 2 ) or arterial oxygen tension (PaO 2 ) are still. V E efficiency can be assessed for carbon dioxide as the slope of the relationship between V E and VCO 2 measured from the beginning of loaded exercise up to the respiratory compensation point, when the slope is linear. Alternatively, V E efficiency can be assessed as the ratio of V E /VCO 2 during exercise, usually at the anaerobic threshold or when it is at its lowest. And what about oxygen? Notably, the V E /VCO 2 and V E /VO 2 ratios are usually plotted in combination, and their changes are similar with a similar nadir, although the V E /VO 2 ratio increases at the anaerobic threshold, while the V E /VCO 2 ratio increases at the respiratory compensation point. Indeed, as opposed to a linear increase of VO 2 during a progressively increasing workload exercise, the V E versus VO 2 relationship is linear only up to the anaerobic threshold, i.e. for a small part of the exercise. Therefore, the slope of the V E versus VO 2 relationship can be calculated as a linear equation only at the beginning of exercise, when V E is more erratic and more affected by psychological stress. In the late 1990s, Baba et al. 2 suggested that the equation be linearised by applying the logarithmic regression, but this implies that the curve is exponential. Is that true? Probably not, because V E in an exercise ramp protocol from rest to peak is progressively under different domains -VO 2 , VCO 2 , PH and temperature, respectively. Notably, all of these domains are characterised by specific V E kinetics. Indeed, we identify thresholds on V E , such as the anaerobic threshold and the respiratory compensation point, both undefinable in an exponential equation. 3 However, the true question is: does OUES tell something different from V E versus VCO 2 ? Indeed, VO 2 includes three major independent diffusion steps limiting oxygen transport: lungs, blood, and capillary to mitochondria. But almost the same applies to carbon dioxide, albeit in a reverse order: muscles to capillary, blood and lungs. However, the pressure gradients of oxygen and carbon dioxide between upper airways and mitochondria are different, as are their degrees of resistance to gas flow. Figure 2 shows that pO 2 decreases from 160 mmHg in inspired air to 100 mmHg in the alveoli and almost to zero in the mitochondria, where it may even become negative in some circumstances, such as during heavy exercise. 4 The pressure gradient for carbon dioxide is much lower, from approximately 45 mmHg in the muscles to 40 mmHg at the end of expiration. The reduction of pO 2 has two quantitatively different steps, the smallest inside the lung and the largest from the capillary to the mitochondria. It is different for pCO 2 changes, which are almost negligible, from the muscles, where carbon dioxide is stored, to venous blood, and from venous blood to expired air. As a matter of fact, end-tidal carbon dioxide is mainly an index of dead space V E , and therefore of V E efficiency. In conclusion, V E versus VCO 2 is mainly lung efficiency, while V E versus VO 2 is a composite of lung, circulation and capillary-to-mitochondria resistance to oxygen flow, i.e. total body efficiency, and not only peripheral efficiency as stated in one of the OUES definitions reported above. Moreover, if OUES is an index of total body VO 2 efficiency, it is unclear to us why we should plot VO 2 versus V E and not versus workload, as in the VO 2 versus workload relationship, because V E is only one of the diffusing limiting steps for VO 2 .
Another major issue is how to 'normalise' efficiency. Indeed, the slope of the V E versus VCO 2 relationship is considered normal below 34, and this cut-off is reported in several guidelines and expert consensus documents. 1,3,[5][6][7][8][9] Indeed, the value of 34 has been found in a population of healthy individuals as the average þ2 standard deviations (SD). Nobody, as far as we know, uses any sort of predicted value for the V E versus VCO 2 slope. However, the V E versus VCO 2 relationship is higher in women than in men, and it increases slowly but progressively with age. 10 Notably, in the present issue of the European Journal of Preventive Cardiology, Hossri 11 proposed OUES reference values for children and adolescents, and a sort of 'normalisation' considering OUES/kg or body surface area (BSA). This is one of the first efforts in the process of V E efficiency normalisation, but are kg or BSA enough? In other words, are any gender differences or any kind of differences related to the development of sexual hormones? And what about V E versus VCO 2 in this young population? Do OUES and the V E versus VCO 2 slope have similar behaviours with age increase in children and adolescents? In our opinion, the paper by Hossri is the first effort in the direction of a normalisation of efficiency, but a more comprehensive assessment including gender, height, race, abdominal circumference, etc. is needed for a proper evaluation.
Regardless of this, Hossri et al. 11 showed that ventilatory requirements are increased in children and adolescents with congenital heart disease either in absolute values or normalised for weight or BSA, this information is clinically relevant and helps to understand the exercise limitation in these patients.
In conclusion, it is important to normalise every measurement, so that we can compare subjects of different age, gender, race, height and weight, but before normalising OUES we need to know what OUES really means. In other words: efficiency of what?

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.