Oxygen needs
Pulmonary and alveolar ventilation
Alveolar gas equation
End-expiratory and arterial PCO2
Coupling between metabolism and ventilation



A very general expression of the metabolic pathways involving oxygen (aerobic metabolism) can be represented in the form of the reaction
                                                         O2 + food water + carbon dioxide + energy                                                   (eq.1)

the stoichiometry of which depends on the type of food substrate utilised. It is important to realize that the above reaction does not mean that O2 interacts as a substrate; in fact, O2 never reacts directly with amino acids, lipids or sugars, nor with their simplified molecular constituents. It simply acts as a proton (and electron) acceptor at the end of the electrons transport chain, which is a complex of enzymes packaged together within the inner membrane of the mitochondria.

 Indeed, the very same properties that make oxygen an ideal electron acceptor would cause major cellular damage if oxygen was left free to react within the cytosol. Oxygen is a major toxic agent, and most animals, including man, would seriously suffer if exposed to high oxygen concentrations for protracted periods of time, and die within a few days if forced to breathe 100% O2 [Fig.1], with dyspnea and hypoxemia resulting from pulmonary edema. If pure O2 was administered at higher pressures, death could be extremely rapid, with convulsions due to the toxic effect of O2 on the nervous system.

Fig.1.   Survival of rats of different age, when exposed to >95% O2 breathing for 3 days.
(Modified from Frank, Fed.Proceed. 44:2328-2334, 1985)

Most cells are equipped with a variety of mechanisms to deal with the oxidative effects produced by internal 'leaks' of oxygen and its radicals; some of these mechanisms are represented by a battery of enzymes (scavengers) designed to immediately neutralize the oxidative effects of the free radicals, whereas other mechanisms are designed to continuously repair the unavoidable structural damages. Superoxide dismutases (SODs) catalyze superoxide (O2.-) into hydrogen peroxide (H2O2); H2O2 is potentially very toxic, especially because it can be reduced in the highly reactive and damaging hydroxyl radical OH-. Normally, this transformation is avoided by cellular catalases and peroxidases, which dismute and reduce H2O2 to H2O. Dismutases, catalases and peroxidases are probably the main cellular players in the defensive team against O2 free radicals. Many others are know, like a-tocopherol and carotenoids (vitamin A), ascorbate (vitamin C), etc.

Hence, one could look at the interaction of O2 with the living matter as a form of fine balance between the advantages coming form its use in the process of aerobic energy production (oxidative phosphorylation) and the risks involved with its handling at the cellular level (oxygen toxicity).

                                                                                                                   Cellular damage (toxicity)  
                                        Oxygen               Interaction with electrons  {  
                                                                                                                   Coupling with ATP-synthase  
                                                                                                                              (Oxidative phosphorylation)  




    In steady state conditions, the question of how much oxygen we need to consume (oxygen consumption, VO2), either to rest, to exercise, or simply to survive (basal oxygen consumption) can be easily answered by direct measurements (e.g. a modified spirometer) or indirect estimates (indirect calorimetry). In steady state, which is the situation here assumed for the sake of simplicity, the uptake of O2 at the pulmonary level corresponds to VO2, i.e. pulmonary uptake = body (tissue) consumption. Equally, the pulmonary elimination of CO2 corresponds to the body production of CO2 (VCO2). In dynamic conditions, however, there may be major differences between pulmonary O2 uptake or CO2 elimination and, respectively, VO2 or VCO2, because of the body stores of O2 (rather small) and of CO2 (very large).

In the adult man, basal VO2 is about 250-300 ml/min, or 3-4 ml/kg/min, and it can increase more than 20 fold during strenuous exercise, until its maximal value, VO2max [Fig.2]. VCO2 is between 70 and 100% of VO2. Values of VO2 and VO2max differ

Fig.2.   Oxygen consumption (O2 ) as function of time during exercise of progressively greater intensity (1<2<3<...), all beginning from rest. Exercise 6 is "supra-maximal", i.e. it does not raise VO2 above what attained with exercise level 5. In the right panel, the VO2s attained at the various exercise levels are plotted against the intensity of the exercise. The plateau in the Power-VO2 function indicates VO2max.

between genders, and with age [Fig.3]; to some extent, these difference persist even after normalization by body weight. Major differences in VO2/kg occur among species, larger mammals needing less O2, per unit weight, than smaller species do.

Fig.3.   VO2max vs age in males and females. After puberty, females have lower values than males, and in both genders VO2max decreases with age. Obviously, for an aerobic exercise, an high VO2max is crucial. However, what would you consider more important, VO2max in absolute values, or VO2max/kg?

Although a small component of O2 can be used in the cytosol by oxidases and oxygenases, almost all VO2 is for the purpose of producing chemical energy and heat. The former is eventually used to drive enzymatic reactions which are energetically unfavourable, to maintain electrochemical gradients, or converted into other forms of energy, including external work and heat. In invertebrates and lower vertebrates (fish, amphibia and reptiles) ambient temperature (Ta) is an important determinant of the speed of enzymatic reaction, and therefore of VO2. As Ta increases VO2 increases, approximately doubling for every 10oC [Fig.4]. Indeed, all these animals are poikilotherms, meaning that their body temperature (Tb) is not constant but varies with


Fig.4.  Changes in the speed of enzymatic reaction (and therefore in VO2, in a steady state aerobic condition) with changes in temperature (Q10 effect). The dashed vertical lines indicate the most common range of body temperature in adult mammals; however, in many cases (e.g. hibernation, torpor, or hypoxia especially in newborns) body temperature can decrease to very low values.

the environment, and their VO2 is Ta-dependent. In homeotherms (birds and mammals), on the other hand, Tb tends to be controlled by mechanisms of heat production (thermogenesis) and heat dissipation. Hence, some of the resting VO2 is for the production of thermal energy, in an effort to maintain Tb within narrow limits, irrespective of changes in Ta. In most mammals Tb is maintained around 36-37oC, whereas in birds is around 38-39oC. It follows, therefore, that in homeotherms, differently from poikilotherms, VO2 increases not only with an increase in Ta, but also when Ta decreases [Fig.5].

Fig.5.   As ambient temperature (Ta) decreases below thermoneutrality, oxygen consumption (VO2) increases, maintaining body temperature (Tb). When thermogenesis does not suffice, Tb begins to fall (critical Ta). The ability to maintain a thermoneutral range is mostly due to an increase in heat dissipation. Eventually, with further increases in Ta, heat-loss mechanisms will not prevent a rise in Tb, which will also lead to a rise in VO2.

Over some range of Ta, VO2 is actually steady at a minimum level, since Tb is controlled by altering the magnitude of heat dissipation. This range is called thermoneutrality, physiologically defined as the range in Ta with maintenance of Tb with minimal normoxic VO2. The specification 'normoxic' is important, because hypoxia can decrease VO2. Above thermoneutrality VO2 increases, partly because of the Q10, partly because of the cost of heat dissipating mechanisms. Below thermoneutrality, VO2 increases because of thermogenesis. The latter includes behavioural thermogenesis (e.g. changes in location, postural changes), shivering (muscle contraction for the purpose of generating heat) and non-shivering thermogenesis (heat produced by specialised organs, namely the brown adipose tissue). The relative contribution of these thermogenic mechanisms, and their threshold of activation, vary among species and with age. For example, in the neonatal period shivering is rare, whereas brown fat thermogenesis represents the major mechanism of heat production.. At very low Ta, VO2 begins to fall, partly because of the Q10 effect, partly because of the inhibitory effect of hypothermia on thermogenic mechanisms.
    Hence, in very cold conditions, thermogenic mechanisms do not suffice, Tb begins to falls and its drop further reducesVO2, leading to a vicious cycle of progressive hypothermia. Thermoneutrality is an important physiological concept, useful also for the interpretation of the interaction between heat producing and heat dissipating mechanisms. However,  it needs to be emphasized that the large majority of homeotherms, even during the early phases of postnatal development, do not live in thermoneutral conditions; their preferred Ta is at the lower end of, or frequently slightly below, thermoneutrality.



Because in most mammals, including newborns the skin represents an irrelevant route for gas exchange, gaseous metabolism, i.e. the exchange of oxygen and carbon dioxide with the environment, must depend on pulmonary ventilation. Hence,


                                                              VO2 = VE ( [O2]I - [O2]E ),                                                      (eq.2)
                                                             VCO2 = VE ( [CO2]I - [CO2]E ),                                                (eq.3)

where I and E represent, respectively, the mean inspired and expired concentrations of O2 and CO2, and inspired and expired ventilation are assumed to be the same*.

[*Strictly, this would be true only if the respiratory exchange ratio (VCO2/VO2) was equal to unity; more often it is between 0.7 and 1]

Hence, changes in VO2 can be accommodated by changing the extraction of O2 from the inspired air, i.e. by widening the I-E [O2] difference, or by increasing VE, in any combination. In addition, a particular level of VE can be achieved by various combinations of breathing frequency (f) and tidal volume (VT).

    The possibility of varying f and VT, i.e. of changing breathing pattern, for any particular level of VE, has profound implications on the ventilation of the alveolar regions (VA), and is therefore crucial for gas exchange and its proper matching to VO2. In fact, VT is the sum of the alveolar volume (VA) and the dead space (VD), and the latter can be taken almost as a constant, since its volume largely reflects the structural characteristics of the airways*.

[*VD increases during inspiration, especially during deep breathing, lowering airways resistance. However, its volume
changes are small compared to the increas in VT]

The following example, from a human subject with VD=200 ml, illustrates this point:

                            V        =         VT                      f ,            and              VA                         f           =                VA

                        (ml/min)                (ml)             (breaths/min)                       (ml)                 (br./min)                   (ml/min)

case a.               4000       =         500                     8                         (500-200)                 8           =             2400

case b.               4000       =         250                   16                         (250-200)                16          =               800


    Hence, for the same VE of 4000 ml/min, the rapid and shallow pattern (case b) reduces VA from 2400 ml/min of case a to only 800 ml/min. Indeed, for any given VE, the deeper the breathing pattern the more efficient it becomes from the view point of alveolar convection, since VD reduces its burden on VT. Eventually, energetic constraints pose a limit to the gas exchange advantage of the deep and slow breathing pattern.


From the above, it is apparent that VE is a parameter of importance in considering various aspects of the energetics and of the regulation of the breathing act; however, from the view point of gas exchange and of accommodating the metabolic needs of the organism VA, not VE, is the parameter of major functional interest. Hence, if we are concerned about the coupling between pulmonary convective mechanisms and metabolic needs, eq 1. should be more properly expressed with respect to the alveolar level (A), in the form of


                                                                        VO2 = VA [O2]insp - [O2]A                                                            (eq.4)


                                                                      VCO2 = VA [CO2]A - [CO2]insp                                                      (eq.5)

Because the partial pressure P of a gas within a gas mixture equals the product of its fractional concentration and the total dry pressure, equations 4 and 5 can be expressed with reference to the partial pressure of O2 (PO2) and CO2 (PCO2), rather than their concentrations:

                                                                        PAO2 = PIO2 - (VO2 / VA) Pb                                                          (eq.6)


                                                                    PACO2 = PICO2 + (VCO2 / VA) Pb                                                       (eq.7)

Pb representing the dry barometric pressure, and neglecting the small difference between inspired and expired ventilation introduced by a respiratory exchange ratio less than unity.
Fig.6 provides the graphical representation of the alveolar gas equations for O2 and CO2 (eq.6 and 7), during air breathing at

Fig.6.   Graphical representation of the alveolar gas equation for O2 (top) and CO2 (bottom), during air breathing at sea level, and at arbitrarily chosen metabolic rates (VO2 and VCO2) of 100 ml/min (continuous line) and 200 ml/min (dashed line). Inspired O 21%, inspired CO2 0%, barometric pressure 760 mm Hg.

sea-level, for two arbitrarily chosen levels of metabolic rate, m (continuous lines) and 2m (dashed lines). In examining these hyperbolic relations, several aspects need to be emphasised.

As VA increases, PAO2 rises toward the inspired value, and PACO2 decreases also approaching the inspired value which, in the usual case of breathing air, is almost zero.

Doubling VA halves the PACO2 value (eq.6); therefore, the absolute drop in PACO2 progressively decreases as VA increases. For example, for an increase in VA from 3 to 6 l/min, PACO2 drops from 44 to 22mm Hg, i.e. by 22 mm Hg. On the other hand, the same increase in VA from 12 to 15 l/min will reduce PACO2 by less than 3 mm Hg.

It is also worth noticing that even a large change in VA when VA is already at high levels has little consequences on PAO2, whereas a small further drop when VA is already low can cause an important fall in PAO2.

The shape and position of the metabolic hyperbola depend on the metabolic level and the inspired concentration of the gases. For example, for CO2 [Fig.6, bottom panel], an increase in metabolic rate, or re-breathing (i.e. breathing air with CO2 in it) will displace the curve upwards. Therefore, the same PACO2 can occur at many different levels of VCO2 and VA combinations. It follows not only that there is no unique relationship between VA and PACO2, but also that an increase in VA does not necessarily imply a decrease in PACO2. It is important to distinguish between an increase in VA relative to its 'normal' reference value, and an increase in VA relative to the corresponding metabolic level. The former is called hyperpnea, the latter hyperventilation.

Hence, hyperventilation can be defined as any situation that results in a drop in PACO2, irrespective of the absolute level of VE and VA (see eq.7). In other words, hyperventilation can occur even with a decrease in the absolute level of VA (hypopnea) if metabolism decreased even more than VA did; this seemingly unusual combination of hypopnea and hyperventilation is not a rare event during acute hypoxia in the neonatal period. Conversely, constancy in PACO2 (isocapnia) indicates that the level of VA is maintaining its proportionality with the metabolic demands (VA/VCO2 = constant); in such a case, the subject is normo-ventilating, again, irrespective of the absolute level of VE or VA.



Let’s consider the signal of a rapid CO2 analyser, sampling the expired air at the mouth. At first, the CO2 concentration (and PCO2) would be close to nil , since the first air expired originates from the dead space, which is filled by the inspired gas (about 0.03% CO2). Only after the whole VD has been expired, PCO2 will increase, to a value that could approach the alveolar value [Fig.7].

Fig.7.  Time profile of alveolar and mouth PCO2 in expiration and inspiration during resting breathing. Mouth PCO2 follows closely the alveolar value toward the end of expiration (end-tidal PCO2)

The time average of mouth PCO2 is the mean expiratory value, whereas the end-expiratory PCO2 (also called end-tidal PCO2) corresponds to PACO2. As indicated in the figure, PACO2 continues to rise until fresh inspired air reaches the alveoli, hence it rises for the whole expiratory phase and for the first part of inspiration (VD washout). The end-tidal PCO2 is a convenient measurement because the alveolar value, in an healthy individual, is very close to the arterial value; in fact, for most practical purposes, PACO2 = PaCO2. Only in the presence of large shunts or major ventilation-perfusion inequalities can PaCO2 be appreciably higher than PACO2.



    Some of the most studied conditions of changes in metabolic level are those during muscle exercise and increased thermogenesis. In several species, including man, VE increases in proportion with VO2 during modest and moderate exercise, and only at the most severe levels of exercise VE increases disproportionately more than VO2 [Fig.8,left]. The latter reflects the additional stimulus on ventilation by the lactic acidosis present at the highest aerobic working levels.

Fig.8.   Examples of the changes in ventilation (VE) when oxygen consumption (VO2) increases because of muscle exercise (left) or cold-induced thermogenesis (right). The constancy in PaCO2 indicates that there is no hyperventilation, i.e. the hyperpnea is proportional to the increased metabolic requirements.

Cold induced thermogenesis [Fig.8, right], circadian oscillations in metabolic rate, pharmacological stimulants of VO2, are other examples of the tight coupling between VE and VO2. In all these cases, in fact, even large changes in VO2 can occur with no or minimal changes in PaCO2.