The Oxygen Cascade During Exercise in Health and Disease

Abstract

The oxygen transport cascade describes the physiological ‘steps’ that bring atmospheric oxygen into the body where it is delivered and consumed by metabolically active tissue. As such, the oxygen cascade is fundamental to our understanding of exercise in health and disease. Our narrative review will highlight each step of the oxygen transport cascade from inspiration of atmospheric oxygen down to mitochondrial consumption in both healthy active males and females along with clinical conditions. We will focus on how different ‘steps’ interact along with principles of homeostasis, physiological redundancies and adaptation. In particular we highlight some of the parallels between elite athletes and clinical conditions in terms of the oxygen cascade.

1. Introduction

In the era of molecular biology it is easy to overlook the central role that the oxygen molecule plays in life. In complex organisms, where most cells are anatomically remote from atmospheric oxygen, a transport and gas exchange system is required to move oxygen from the air to the tissues and supports the continuous generation of ATP via oxidative metabolism. To defend whole body homeostasis, this system must be robust enough to sustain vast increases in oxidative metabolism during exercise, survive at high altitude, and – at the opposite end of the homeostatic spectrum – withstand substantial insults to key elements of the system associated with disease. In this review we will follow the movement of oxygen from the air to the tissues and use examples ranging from elite athletes, rare patients, and comparative biology to illustrate key principles of oxygen transport in humans. The impressive adaptive nature of the cardiopulmonary system will be emphasized and the redundant nature of physiological control mechanisms and related anatomical design features will be highlighted.

2. The Problem

The partial pressure of oxygen in air at sea level is ~150 mmHg but in the mitochondria of exercising skeletal muscle the partial pressure of oxygen can be ~100-fold lower without significant engagement of anaerobic energy metabolism. Thus, the question is how does this happen and what systems are engaged as it occurs? In this context, it is also important to consider the range of oxygen consumption that can be seen in humans. In healthy young adults, resting metabolic rate is approximately 3.5 mL O₂ · kg⁻¹ of body weight · min⁻¹ and it can increase to over 90 mL O₂ · kg⁻¹ · min⁻¹ in at least some of the most aerobically-trained elite athletes.¹ At the opposite end of the O₂ uptake spectrum, even modest metabolic rates require physiological redundancies to be engaged in patients with diseases that hinder the ability to take up and transfer oxygen. Notable examples include marked tachypnea or tachycardia even during very modest levels of exercise in patients with diseases, like chronic obstructive pulmonary disease or congestive heart failure, along with a marked redistribution of blood flow away from inactive tissues to the active muscles.²

3. The O₂ Transport cascade

The oxygen transport cascade describes the path of oxygen from the atmosphere to the tissues.³ In general, it is an anatomical map starting with the extra and intrathoracic large airways and ending in the mitochondria. Beyond the lungs and alveolar pulmonary capillary membranes, the right and left heart are connected in series and pumps blood across the interface of the pulmonary capillaries and alveoli. Oxygenated blood then leaves the left ventricle and via the systemic circulation oxygen is distributed to the microcirculation in the peripheral organs for use by the tissues. The venous system then returns the deoxygenated blood back to the right heart and the cycle continues. Figure 1 is a schematic of the cascade and shows the major steps. The overall behavior of the cascade can also be described via the Fick equation which simply states oxygen consumption (V̇O₂) = cardiac output (Q̇) × arteriovenous oxygen difference (a-vO₂diff).

Figure 1.

Schematic of the oxygen cascade during exercise.

$${\dot{V}}_{O_2}=Q\left(C_{a{O}_2}-C_{v{O}_2}\right)$$

In this context, elements of the cascade contribute to the variables in this equation. It is also important to note that the magnitude of oxygen consumption that can be supported by the cascade is vast. As noted above, resting oxygen consumption is 3.5 mL O₂ · kg⁻¹ · min⁻¹ or 1 metabolic equivalent (MET). A four MET peak exercise capacity (~14 mL O₂ · kg⁻¹ · min⁻¹) is generally thought to be required for most humans to successfully engage in activities of daily living without limitation.⁴ During heavy exercise in untrained non-athletic humans oxygen consumption can increase to 10–12 METS. Many young, trained healthy humans are able to sustain 15 METs with levels as high as 20–25 METs seen in elite endurance athletes.⁵

In the next section, we will begin to discuss the anatomical features and mechanisms responsible for this oxygen transport, how such a huge range is possible, and how it can be influenced by exercise training to selected pathophysiologic changes in the cardiopulmonary system. We will also describe examples of how each step of the oxygen cascade can be limited. A useful reference is Figure 2 which shows the ‘typical’ response to exercise of many variables that we subsequently detail. Note this Figure shows relative change from rest rather than absolute values. This was done because absolute values can vary considerably depending on age, sex, height, training status, whereas the relative change is more consistent (but exceptions do exist as noted below).

Figure 2.

The respiratory, cardiovascular and metabolic response to aerobic exercise. Each x-axis represents progressive exercise starting at rest and ending at maximal intensity. The y-axis are unitless and are intended to show the relative change. Abbreviations: PAO₂, alveolar oxygen tension; A-aDO₂, alveolar-to-arterial oxygen gradient; PaO₂, arterial oxygen tension; PaCO₂, arterial carbon dioxide tension; SaO₂, oxyhemoglobin saturation; V̇_E, minute ventilation; V_A, alveolar ventilation; V_T, tidal volume; Fb, breathing frequency; TPR, total peripheral resistance; SBP, systolic blood pressure; MAP, mean arterial pressure; DBP, diastolic blood pressure; MSNA, muscle sympathetic nerve activity; Q, cardiac output; SV, stroke volume; HR, heart rate; PvO₂, mixed venous oxygen tension; HCO₃⁻, arterial bicarbonate; K⁺, arterial potassium ion concentration; A-vDO₂, arterial-to-venous oxygen difference; V̇CO₂, carbon dioxide production; V̇O₂, oxygen uptake; Lac, arterial lactate concentration.

4. Alternating Convective and Diffusive Steps

A key feature of the oxygen cascade is that so-called convective and diffusive steps alternate. This means that there is what could be described as bulk transport of oxygen followed by diffusion across a membrane and then uptake by a tissue. For example, minute ventilation is a convective action that moves volumes of oxygen-containing air in and out of the lungs. The composition of the air changes as it is mixed with the gases already in the lungs. However, no gas is exchanged until there is a diffusive step at the interface of the alveolar/pulmonary capillary membrane. There oxygen is transferred to a tissue, the red blood cells, and then distributed via the large blood vessels to the microcirculation in another convective step. Oxygen then leaves the microcirculation by diffusion across the capillary tissue interface. Once in the tissue it is used by the mitochondria for the purposes of oxidative metabolism. As noted above this means that the partial pressure of oxygen drops from around 150 mm of mercury in inhaled air to potentially very low pressures in the tissues. Parenthetically, mitochondrial pO₂ values exercising skeletal muscles may be less than 1 mmHg.⁶

Due to the fact that the oxygen molecule is nonpolar, its solubility in water is poor. This leads to unique challenges for oxygen transport. With regard to convective transport, oxygen’s low solubility means that plasma itself only carries ~0.5 vol% of oxygen when equilibrated with atmospheric pressure. This minimal amount of dissolved oxygen would render even modest changes in oxygen demand unattainable. For example, a V̇O₂ of 3 L · min⁻¹ would require a cardiac output of 660 L · · min⁻¹ without the hemoglobin in erythrocytes as an oxygen carrier. With regard to diffusion, the low solubility of oxygen in tissue means that the low concentrations of oxygen necessitate short diffusion distances to achieve the gradients required to supply the mitochondria in working muscle. In particular, if the oxygen consumption in muscle performing heavy exercise is 35 ml O₂ · 100cm³ tissue⁻¹ · min⁻¹ (about a 100 fold increase above resting), the diffusion distance can be calculated to be at most 40 μm, meaning that oxygen must be brought by convection to within this distance of every point in the tissue. Thus a dense mesh of capillaries in the microcirculation is needed to achieve this condition. The observed fiber diameters of ~50 μm in skeletal muscle are consistent with this estimate of the maximum diffusion distance to sustain activity.

The low solubility of oxygen in plasma and alveolar tissue also limits diffusive exchange in the lung. However, the very large surface area for gas exchange in the healthy lung, combined with a very short diffusion distance between alveoli and blood (~2 μm) provides adequate diffusive flux.^7,8

5. Ventilation

The process of moving air in and out of the lungs to deliver oxygen to the alveoli and remove carbon dioxide is known as pulmonary ventilation. At rest in healthy young humans, minute ventilation is 5 to 10 liters per minute with breathing frequencies of around 10 breaths per minute and tidal volumes of 0.5–1.0 L. The range is principally dictated by height/weight but a litany of other factors can influence this (e.g. hormones). The diaphragm is the primary muscle involved in moving air at rest. During exercise or in the presence of excessive respiratory muscle work, other respiratory muscles (e.g. intercostals, sternocleidomastoid) become progressively more active, along with the diaphragm. Due to the reciprocating (in-out) nature of breathing in humans, gas is not exchanged in the large airways and is mixed with alveolar gas. This “dead space” in the large airways along with the reciprocating nature of ventilation explains the drop in oxygen partial pressure from the atmosphere to the alveolar gas. It also explains why alveolar ventilation is always less than minute ventilation.

Ventilation is driven by complex neural circuits located in the brainstem that include a central pattern generator and chemosensor along with feedback from peripheral sensory nerves located in the carotid bodies, aortic arch, muscles and lungs.^9,10 Thus, there is both feed-forward and feedback regulation of ventilation.¹¹ During rest at low altitude the primary feedback mechanisms that control minute ventilation are related to CO₂ and pH. Hypoxia caused by either high altitude or disease can engage the carotid chemoreceptors which are powerful drivers of increased minute ventilation.^12,13

Each anatomical and neural element involved with ventilation and its regulation can be subject to one or more disease processes. For example many forms of lung disease increase dead space so that with each breath there is less gas exchange in the alveoli and more so-called ‘wasted’ or dead space ventilation. There can also be neurological and neuromuscular diseases that affect brainstem control of breathing and or diaphragm function. Furthermore, the behavior of peripheral chemoreceptors that sense CO₂, pH and O₂ is altered in many disease processes. Perhaps the most dramatic of these syndromes is congenital central hyperventilation syndrome (Ondine’s curse), which requires the patient to engage in conscious efforts to breathe even at rest.

The temporal pattern of each breath is also important to consider. At rest inspiration takes perhaps a second or more and is accomplished almost exclusively by contraction of the diaphragm, whereas expiration at rest is passive and occurs due to the elastic recoil properties of the lung and chest wall.¹⁴ During periods of increased ventilation, this pattern changes: inspiratory time shortens, additional inspiratory respiratory muscles are recruited to generate larger tidal volumes during exercise, and expiration can become active and aided by the muscles of the chest.

At rest, the tidal volume of each breath is roughly 500 mL. About 150 to 200 mL of this breath represents dead space with 300 to 350 mL of alveolar ventilation available for gas exchange. In health, most of the dead space is anatomical in nature and remains unchanged throughout exercise. Each of these resting breaths is roughly 10% of vital capacity. Typically, as exercise intensity increases and necessitates greater ventilation, it is tidal volume that increases first and plateaus at ~40–60% of vital capacity and ~60% of peak exercise capacity. This increase in tidal volume is achieved by utilizing both inspiratory and expiratory reserve volume and optimizes alveolar ventilation and minimizes mechanical work.¹⁵ Alveolar ventilation is optimized because anatomical dead space is fixed and increasing tidal volume minimizes the dead space to tidal volume ratio thus improving alveolar ventilation. Mechanical work is minimized because there is an ideal balance between minimizing elastic (via appropriate tidal volumes) and resistive (via lower flows and frequencies) respiratory work for any given ventilation. Once tidal volume plateaus, greater breathing frequency plays a greater role in increasing total ventilation. During exercise it is not uncommon for minute ventilation to rise to values of 90–100 liters per minute or more in healthy young subjects (Figure 3). A similar breathing pattern is observed in elite athletes, except that higher end-exercise values in excess of 150 liters per minute can be seen, and in very large male rowers values in excess of 240 liters per minute have been reported.¹⁶

Figure 3.

Typical maximal ventilations in different populations of males and females. Rowers (males)⁸⁸ and cross-country skiers (females)⁸⁹ represent highly trained individuals with very high ventilations. V̇_E, minute ventilation

How are these impressive volumes of ventilation generated? First, there is an increase in what has been described as “central command”.¹⁷ Central command is a feed-forward mechanism that increases ventilation and heart rate at the onset of exercise in proportion to the effort associated with the exercise. This is essential because “waiting” for feedback from the peripheral chemoreceptors could lead to dangerous changes in pH and increases in CO₂, so a feed-forward control mechanism is required. During the very high volumes of ventilation described above, inspiratory time progressively decreases from a second or more to as little as 0.5 seconds, but more importantly expiration becomes active and expiratory time decreases dramatically. In elite athletes it is not unusual to observe 60 breaths per minute with an inspiratory and expiratory time of 0.5 seconds each and a tidal volume of 2.5 to 3.5 liters per breath. Fine tuning of exercise hyperpnea occurs through feedback from the peripheral chemosensors and also type III, IV sensory afferents in the contracting respiratory and skeletal muscles.^18,19 Interestingly, there is evidence that oscillations in pH/CO₂ may contribute to the hyperpnea of exercise but the primary driver is almost certainly a feed forward mechanism.²⁰

The high breathing frequencies and tidal volumes seen during heavy exercise, especially in elite athletes, bring us to an outstanding example of how similar physiological constraints can meet at the far ends of the physiological spectrum. In this context, the high respiratory frequencies, and consequently flows, and tidal volumes noted above cause some athletes to encroach upon the maximal capacity of their expiratory flow volume loops (Figure 4, blue lines). This means that expiratory flow is constrained by airway anatomy and elastic recoil such that even with additional respiratory muscle effort, higher expiratory flow cannot be attained at a given lung volume. When this occurs during exercise it can lead to relative hypoventilation, exercise-induced hypoxemia and CO₂ retention. This phenomena is also more common in elite female athletes because women tend to have smaller lungs and airways for a given body size.^21,22 Evidence for this is from exercise-induced hypoxemia susceptible athletes where the hypoxemia can be partially reversed by breathing helium-oxygen (heliox) gas mixtures.^23,24 Heliox has a lower density compared to nitrogen balance gas mixtures and increases the propensity towards laminar flow which allows higher flows and higher minute ventilations to be achieved. The greater ventilation can increase partial pressure of oxygen in alveoli and can reduce hypoxemia and CO₂ retention.^23,24

Figure 4.

Ventilatory response to exercise in terms of lung volumes and flow. Leftmost side represents a spirogram showing the change in lung volumes and flows during progressive exercise in a young healthy female (solid black), an aerobically trained young female (solid blue) and a healthy older female (solid red). The dotted lines represent the change in lung volumes. The rightmost side shows a maximal flow-volume curve along with maximal flow-volume loops for the same color coordinated subjects.

In the absence of experimental manipulations, to achieve the higher ventilations and avoid this expiratory flow limitation, some athletes will increase their operating lung volume in order to take advantage of the higher peak flows (Figure 4). While this does allow greater ventilation, it comes at the cost of greater work of breathing and places the diaphragm in a mechanically compromised position.²⁵ Expiratory flow limitation and relative hyperinflation are especially prominent in healthy older individuals during exercise as the normative aging process reduces the maximal flows, yet the metabolic demands still require a considerable ventilation.^26,27 These mechanical constraints are especially prevalent in older women.²⁸ In general ventilatory constraints in youth are normally only seen in more trained individuals because unlike the cardiovascular or musculoskeletal systems, the pulmonary system appears to show minimal to no improvement with physical training.²⁹ This is depicted in Figure 4 whereby the trained athlete and the typical female have similar maximal expiratory flow volume envelopes, but the trained females maximal tidal loop (hashed blue line) is greater. In other words, the increasing demands placed on the respiratory system from physical training are met with a static capacity.³⁰

A similar situation, expiratory flow limitation and/or an increased operating lung volume, is seen in individuals with structural lung disease such as chronic obstructive pulmonary disease, some forms of asthma, and especially in Masters athletes²⁸ (over 60 for the purposes of this discussion) (Figure 4 red lines).²⁷ Note the older healthy individual has reduced expiratory flow which results in flow limitation being common and often severe.²⁸ The key point is that in both elite athletes and in patients the structural factors in the lung which limit peak flow during heavy exercise can limit exercise tolerance or performance.³¹

Along with limitations to ventilation due to structure of the lungs and airways, the respiratory muscles themselves can also impact exercise performance. It is now appreciated that the diaphragm³² and expiratory muscles³³ can fatigue with some forms of exercise and this fatigue impacts exercise performance.³⁴ It is important to note that higher respiratory muscle work in and of itself (i.e. in the absence of exercise) does not appear to elicit diaphragm fatigue,³⁵ rather there appears to be a need for the metabolic changes and potentially blood flow redistribution that occurs with exercise. Fatigue of the respiratory muscles can also influence locomotor muscle fatigue, which will also impact performance.^36,37 On the opposite end of the spectrum, respiratory muscle fatigue is implicated in forms of respiratory failure and may play a role in the ability to wean patients from mechanical ventilation.

Finally, in an era marked by massive searches for genetic explanations underpinning human variation, while there are clearly genetic syndromes like alpha-1 antitrypsin deficiency that can influence pulmonary function, large Genome Wide Association Studies (GWAS) have failed to explain much if any of the population variability in human lung function. In fact, GWAS studies have failed to shed much if any light on any potential DNA based contributions to population wide phenotypic variation in any of the key steps of oxygen transport cascade discussed in this review.³⁸

6. Oxygenation

At rest the partial pressure of oxygen in the venous blood entering the lungs is ~40 mmHg and 75% saturated while the partial pressure of oxygen in the blood leaving the lungs is ~90–100 mmHg and 95–98% saturated. The first step in the cascade - ventilation - has taken oxygen from the air and delivered it to the alveoli. The next step is to oxygenate the blood. For this to occur, oxygen must diffuse across the alveolar pulmonary capillary membrane and bind to hemoglobin in red blood cells. Compared to CO₂ and other gases, oxygen is somewhat less diffusible and it takes some time (about 0.3 seconds) for the oxygen tension in the blood flowing through the lungs to equilibrate with the oxygen levels in the alveolar gas. Additionally, there is a small alveolar to arterial O₂ gradient (1–5 mmHg) even in individuals with absolutely healthy lungs and normal lung function at rest. This mild gas exchange impairment at rest is most likely the result of ventilation-perfusion mismatch or a small amount of shunted blood.³⁹

Once again the factors that can limit diffusion across the alveolar pulmonary capillary membrane in patients and elite athletes during heavy exercise have some parallels. The most notable diagnosis associated with arterial hypoxemia in patients are various forms of interstitial lung disease like idiopathic pulmonary fibrosis. In these diseases, chronic inflammation increases the thickness of the alveolar pulmonary capillary membrane, and this limits the diffusion of oxygen from the alveoli to the blood. These diseases and others, such as chronic obstructive pulmonary disease also commonly result in ventilation-perfusion mismatches that are related to disease severity and the inherent heterogeneity of the disease.^40,41 Other causes of arterial hypoxemia include shunt as seen in various forms of congenital heart disease when some of the venous blood is not pumped through the lungs and is directed back to the left heart. There are also rare hemoglobinopathies and conditions like carbon monoxide poisoning which limit the ability of the red blood cells to bind oxygen once it has diffused across the pulmonary capillary membrane.

The pathophysiological examples highlighted above have some parallels during heavy exercise, especially in elite athletes. That is, some apparently healthy young athletes can develop significant arterial hypoxemia during intense exercise.⁴² The cause of the hypoxemia is multifaceted and variable between individuals, but is principally due to relative alveolar hypoventilation and gas exchange impairments.⁴² Relative alveolar hypoventilation can result from mechanical ventilatory limitations (e.g. flow limitations)^23,25 or inadequate neural drive to breathe.⁴³ However, relative alveolar hypoventilation usually constitutes less than half of the apparent hypoxemia and is rarely the exclusive cause. Rather, all individuals (regardless of aerobic fitness) will develop some degree of gas exchange impairment during exercise.³⁹ The gas exchange impairment occurs through ventilation-perfusion mismatch and diffusion limitation, with the former more pronounced early on in exercise.³⁹ In modestly trained individuals, this increased alveolar to arterial oxygen gradient is offset by an adequate compensatory hyperpnea which ensures that their arterial oxygen pressure remain near resting pressures. In trained athletes however, the extraordinary demand for oxygen results in the alveolar to arterial oxygen gradient being so large (e.g. 25–40+ mmHg) that it is either too excessive to offset and/or mechanically unachievable.²³ Due to small airways and more mechanical constraints, healthy women may be more prone to arterial hypoxemia during exercise, but this is not yet definitive.⁴⁴

Other potential mechanisms for hypoxemia during exercise include high pulmonary artery pressures which can cause subclinical pulmonary edema and congestion that can limit the diffusion of oxygen across the alveolar pulmonary capillary membrane. The transit time of blood across the alveolar pulmonary capillary membrane is also extremely short during heavy exercise and there is potentially not enough time for O₂ equilibration to occur before the blood leaves the capillary.²⁴ Finally, the partial pressure of oxygen returning to the heart during heavy exercise is quite low (PO₂= ~15 mmHg and 25% saturation or less) and in conjunction with the factors noted above further challenge the ability of the oxygen transport across the alveolar pulmonary capillary membrane. This phenomenon can be further amplified by exposing an individual to hypoxia during exercise. The end result of the above in elite athletes is arterial blood gas values that approach what many would consider pathological. The example below is from a healthy 19 year old, female athlete with normal pulmonary function (all values >105% predicted) at maximal exercise (V̇O_2max 3.4 L · min⁻¹, 56 mL · kg⁻¹ · min).²³ At maximal exercise, their arterial PO₂, SaO₂, A-aDO₂, PaCO₂ and pH were: 70 mmHg, 89.5%, 39 mmHg, 39 mmHg and 7.17; respectively. Note how this individual had developed considerable arterial hypoxemia which was due to both a lowered arterial oxygen tension (via relative alveolar hypoventilation and considerable gas exchange impairment) and a right-ward shift in the oxygen dissociation curve. Immediately post-exercise (within minutes) however, her arterial PO₂, SaO₂, A-aDO₂, PaCO₂ and pH were: 99 mmHg, 98.5%, 9 mmHg, 31 mmHg and 7.33; respectively. Importantly, this case and others demonstrate that stopping exercise returns these values back to ‘normal’ almost immediately (and probably within seconds).⁴²

7. Cardiac Output

In the absence of disease, environmental hypoxia and with the exception of heavy exercise in some elite athletes, the blood has now been successfully oxygenated at the alveolar pulmonary capillary interface. It then returns to the left heart where it is pumped to the systemic circulation by the left ventricle. A useful rubric is to think about the conductive step of systemic oxygen delivery as the simple product of cardiac output and arterial oxygen content. The determinants of cardiac output are straightforward, heart rate × stroke volume. Heart rate is determined by the intrinsic cardiac pacemaker activity of the heart (about a hundred beats per minute) along with the activities of the cardiac parasympathetic and sympathetic nerves. The determinants of stroke volume include preload, afterload, and contractility along with the chamber size of the ventricle.

Because oxygen is not very soluble in blood, arterial oxygen content is essentially the product of blood hemoglobin concentration and arterial saturation based on the properties of the O₂-hemoglobin dissociation curve.⁴⁵ At rest cardiac output is roughly 5 liters per minute and arterial oxygen content is roughly 20 mL per hundred mL of blood. This means that roughly one liter of oxygen leaves the left ventricle each minute at rest in a generic healthy young person. These values vary modestly depending on differences in body size and represent estimates and values determined in relatively healthy males weighing around 70kg. In the absence of any disease, females typically have ~10% lower hemoglobin concentrations than males of similar size and stature.

Based on the values above, systemic oxygen delivery at rest is about 1 L per minute and systemic oxygen consumption is on the order of 250 to 350 ml per minute at rest. This means that “excess” oxygen is delivered to the peripheral tissues in resting humans. It also means that cardiac output can decline markedly and still supply sufficient oxygen delivery provided the extraction of oxygen from the arterial blood is able to compensate for any fall in cardiac output. This ability to compensate for a reduction in cardiac output explains why humans are able to survive marked reductions in oxygen delivery during acute conditions like blood loss and when left ventricular function is compromised during chronic forms of heart failure.

During exercise cardiac output can increase roughly four-fold in young, healthy humans.^46,47 This is primarily due to a three-fold increase in heart rate from 60–70 beats per minute at rest to around 200 beats per minute during maximal exercise. This increase in heart rate is accompanied by a 20 or 30% increase in stroke volume. The increase in heart rate occurs because parasympathetic tone to the heart is withdrawn and sympathetic activity to the heart increases dramatically.⁴⁸ Stroke volume increases as a result of increased venous return from the periphery as a result of the skeletal and respiratory muscle pump translocating blood to the central circulation along with increased cardiac contractility caused by activation of the cardiac sympathetic nerves and circulating catecholamines. In healthy young men, the combination of these factors leads to a cardiac output of 20 liters per minute with a systemic oxygen delivery of 4 liters per minute to support an oxygen consumption of about 3 liters per minute. In contrast to the case at rest, a much higher fraction (~75 %) of the oxygen delivered from the heart to the periphery is used by the tissues.

In elite athletes stroke volume can be double that seen in average humans and peak cardiac outputs of 35–40 liters per minute or more have been observed.^49,50 This means that 7–8 liters of oxygen leave the heart each minute to support a systemic oxygen uptake as high as 6 or even 7 liters per minute. In conditions like congestive heart failure the situation is reversed and while heart rate can rise, unless compromised by damage in the cardiac conduction system or drugs, stroke volume is reduced and systemic oxygen delivery is very limited.

Most importantly, there is a linear relationship between peak cardiac output, and peak oxygen consumption ranging from very low values in patients with congestive heart failure or conditions like mitral stenosis to very high values in the elite athletes we have described above. Figure 5 shows this relationship. In the case of all humans, activation of the sympathetic nervous system leads to a redistribution of blood flow away from inactive and relatively over perfused tissues like the kidney and liver so that a very high fraction of cardiac output can be directed to the exercising skeletal muscles. During heavy exercise visceral blood flow can be reduced by 75% compared to values measured during rest.

Figure 5.

The relationship between maximal oxygen uptake and cardiac output and red blood cell volume. V̇O_2max, maximal oxygen uptake; Q_max, maximal cardiac output; RBCV, red blood cell volume

Along with changes in cardiac output, how arterial blood pressure changes with exercise is also critical to the oxygen cascade. With healthy individuals during dynamic exercise there is an intensity dependent rise in systolic blood pressure with minimal (if any) change in diastolic pressure. The result is that mean arterial pressure proportionally rises with progressive exercise, but not to the extent of systolic blood pressure (Figure 2). The rise in systolic pressure is largely due to the enhanced contractility and cardiac output while the maintenance of diastolic blood pressure is due to the progressive fall in total peripheral resistance. This relationship is shown with the physiological equivalent of Ohm’s law whereby mean arterial pressure is the product of cardiac output and total peripheral resistance (MAP= Q × TPR). The ‘controlled’ variable in the preceding equation is mean arterial pressure and its regulation is governed in part by arterial and cardiopulmonary baroreceptors.⁵¹ These baroreceptors continuously monitor blood pressure/volumes and alter autonomic activity to ensure arterial pressure is maintained within a narrow range. For example, in response to increased arterial blood pressure, there is a reduction in sympathetic activity and increase in parasympathetic activity with the end result being a lowering of blood pressure to previous levels.

If baroreceptors protect against fluctuations in blood pressure, how does mean arterial pressure rise with dynamic exercise? Previously it was thought that these baroreceptors were ‘turned off’ during exercise. Rather, it is now appreciated that the arterial baroreceptors are still functional during exercise, but they are ‘reset’ to defend a higher arterial blood pressure.^52–54 The mechanisms responsible for this resetting are a combined (and integrative) function of cardiopulmonary baroreceptors, the exercise pressor response and central command.^52,53 The end result is that as exercise intensity increases, the operating point of the baroreceptors also increases.

8. Red Cells/Hemoglobin

Hemoglobin in red blood cells is almost exclusively responsible for carrying oxygen in blood. This means that anemia or reduced hemoglobin content can potentially compromise oxygen delivery. Likewise increased hemoglobin content can also influence oxygen delivery especially during exercise. Normally hemoglobin content are 12 to 16 grams per dL of blood in healthy young women and men. At rest, anemia is associated with an increase cardiac output so that oxygen delivery remains constant in the face of mild anemia. During severe anemia, cardiac output can double or even triple, and increased oxygen extraction also operates to preserve tissue oxygenation when hemoglobin content are reduced to values of ~ 5 grams per dL.⁵⁵ There is also a right shift in the oxygen hemoglobin dissociation curve during anemia which facilitates the off-loading of oxygen at the tissues. When anemic subjects are not hypoxic, the desaturated venous blood can be re-saturated in the pulmonary capillaries in spite of any right shift in the oxygen hemoglobin dissociation curve.

By contrast, the main issue in patients who are polycythemic with high concentrations of hemoglobin and high hematocrits is increased blood viscosity and the potential for medical complications like stroke arising from it. However, the evidence based guidelines for which patients should undergo therapeutic phlebotomy is limited with the best data coming from a large randomized trial in patients with polycythemia vera.⁵⁶ Additionally in many tissues oxygen delivery is not compromised until hematocrit values exceed ~70%.⁵⁷

In addition to hemoglobin and hematocrit, blood volume is also a determinant of oxygen delivery. At rest, total blood volume influences filling pressures in the heart and as a result can also affect cardiac output. Of particular note is the case of exercise, where endurance exercise training can increase total blood volume and red cell mass, although blood volume appears to increase relatively more. Importantly, total body hemoglobin and red blood cell mass (like cardiac output) are key determinants of maximal exercise capacity (Figure 5). This relationship along with the relationship between cardiac output and peak exercise capacity highlights the central role of oxygen delivery in exercise, especially heavy exercise. It also explains why strategies like blood doping and the use of transfusion or erythropoietin (illegally) increase performance in sport.

The oxy-hemoglobin dissociation curve (ODC) has been mentioned several times in this review. This curve describes the curvilinear relationship between the partial pressure of oxygen and the fraction of hemoglobin that is saturated with oxygen in the blood. The curvilinear nature of the ODC is a result of cooperative interactions of the four hemoglobin chains that make up the quaternary structure of hemoglobin and the way their affinity for oxygen changes as oxygen is loaded and unloaded from them. Important to exercise, the ODC is influenced by pH, temperature and CO₂, and 2–3 bisphosphoglyceric acid, with lower pH, higher temperature and higher CO₂ causing a decrease in affinity or ‘right shift’ and making it easier to unload oxygen at the tissues, especially active skeletal muscle. This property of hemoglobin facilitates oxygen delivery to tissues when demand is high as in exercising muscles, or when there is tissue ischemia. It also takes physiological advantage of the fact that hemoglobin is almost completely saturated at relatively modest partial pressures of oxygen. This means that unless an individual is at high altitude or has some pathophysiologic condition in the lung that reduces alveolar PO₂ or pulmonary diffusion, there is typically adequate partial pressure of oxygen to almost fully oxygenate even right shifted hemoglobin as it passes through pulmonary capillaries. This series of events also underpins the general idea, taught in introductory biochemistry and physiology courses, that a right shift in the ODC is generally protective against hypoxia.

The nature of what constitutes an optimal ODC under varying physiologic circumstances is in fact much more complex than the simple right shift is protective against hypoxia narrative. For example, many animals evolutionarily adapted for life at high altitude have a left shifted ODC. This includes camelids like the llama and alpaca, and most notably birds that excel at high altitude migratory flight like the bar-headed goose that traverses the Himalayan Mountains.⁵⁸ Under these circumstances the physiological strategy is to load more oxygen at the lung and allow low oxygen partial pressures at the tissue to facilitate offloading. This offloading is also enhanced in some species by small red blood cells and increased capillary density.⁵⁹

In this context, there are strains of Dorset sheep that contain either left or right shifted hemoglobin. At rest, sheep with left shifted hemoglobin are much more resistant to severe hypoxia, while animals with right-shifted hemoglobin curves are more resistant to severe anemia.⁶⁰ Along these lines, our laboratory has recently studied humans with rare hemoglobin variants that are left shifted. These subjects show preserved capacity and gas exchange during experimental hypoxia.⁶¹ Thus, the extent to which a right or a left-shift in the ODC is physiologically advantageous depends on the physiological challenge to which the organism is exposed. The findings from evolutionary biology and our recent observations in rare patients also point to the limits of extrapolating molecular interactions of proteins like hemoglobin to complex whole body responses to acute exercise in hypoxic environments. It is also interesting to speculate about the potential development of drugs that right shift the ODC to treat forms of acute and chronic hypoxia in patients, or left-shift the ODC to improve pulmonary oxygenation in patients with conditions like acute respiratory distress syndrome.

9. Distribution of Blood Flow

Once the oxygen rich blood leaves the heart, it must be directed to the appropriate tissue in relation to their metabolic activity. At rest, the majority of the 5 L of cardiac output is directed to the brain, heart and visceral organs (e.g. liver, kidney, gastrointestinal tract), and oxygen extraction in the visceral organs is low. During exercise, the skeletal muscle metabolic rate increases and so too does its relative share of blood flow. That is, cardiac output is distributed in proportion to oxygen uptake of the tissue. So much so, that at maximal exercise in a young healthy individual, nearly 85% of cardiac output is directed to active skeletal muscles while blood flow to the heart increases, cerebral blood flow is maintained or increases slightly^60,62 and visceral blood flow is reduced dramatically.

This redistribution of blood flow occurs as a result of local vasodilation in the contracting muscle coupled with increased global sympathetic activity. That the increased sympathetic activity does not result in vasoconstriction in the active muscles is termed “functional sympatholysis⁶³, and is an important mechanisms governing exercise hyperemia.⁶⁴ Interestingly, maximal muscle blood flow conductance is restrained during maximal exercise.⁶⁵ That is, when exercised in isolation (i.e. single leg), maximal blood flow and vascular conductance is greater compared to when multiple muscle groups (i.e. legs and arms) are active. It is proposed that this vascular constraint is to preserve arterial blood pressure because the maximal dilatory capacity would outstrip the cardiac output’s ability to compensate. In fact, studies in the 1960s showed that patients who had been treated for severe hypertension with surgical sympathectomy, actually had a drop in blood pressure while exercising in the supine position.⁶⁶ This sympathetic restraint of maximal conductance generally only occurs at or around maximal exercise in healthy subjects; however, in clinical populations, (e.g. chronic heart failure), this redistribution of blood flow during exercise occurs at much lower intensities (~60% peak work). An analogous situation is the mismatch of vasodilation and cardiac output that can been seen in a septic patient.⁶⁵ In such patients cardiac output can be high and there is marked peripheral vasodilation that has a limited vasoconstrictor response to catecholamines. During exercise vasodilation with limited vasoconstrictor responses in skeletal muscle drives blood flow to the contracting muscles while preserving blood pressure.^65,67 In sepsis, the same responses “inappropriately” drive blood flow to tissues out of proportion to their demand for oxygen and threaten blood pressure.

While not all skeletal muscle can be maximally perfused during intense exercise, there are indications of a blood flow hierarchy.^67,68 Specifically, the respiratory muscles appear to receive adequate blood flow at the expense of other skeletal muscles via sympathetically mediated redistribution.^68–70 That is, based on estimates of the oxygen cost of breathing, the respiratory muscle will command 10–15% of total cardiac output,^69,70 but this may come at the expense of other active skeletal muscles. The selective distribution was demonstrated by experiments showing that when respiratory muscle work was increased so too did respiratory muscle blood flow and this was at the expense of the active skeletal muscles.^71–73 This is an excellent example of how a vital function (i.e. breathing) is defended at the expense of other body systems.

A complex network of arterioles, capillaries, and venules is responsible for adequate delivery of oxygen at the tissue level.⁷⁴ Because of the network structure and spatial and temporal variations (heterogeneity) in oxygen demand, the microcirculation adjusts flow and oxygen delivery by actively varying vessel diameters in a process termed blood flow regulation to maintain an adequate tissue PO₂.⁷⁵ Since capillaries lack smooth muscle and have a limited ability to control their diameter, hypoxia detected at the capillary level generates a signal that is conducted upstream to feeding arterioles that vasodilate, resulting in increased flow and oxygen delivery.^76,77 If hypoxia persists, long term changes in oxygen pressures lead to remodeling of the vascular network, which includes structural adaptation of existing vessels and formation of new vessels (angiogenesis).⁷⁸

At the level of the microcirculation, the particulate nature of blood also becomes relevant. In small vessels such as capillaries, the blood cells tend to migrate toward the center, surrounded by a plasma sleeve; because the overall velocity of the blood is higher than the velocity of the blood cells, the tube hematocrit in these small vessels is lower than the overall discharge hematocrit (Fahraeus effect).⁷⁹ Because of this segregation, the viscosity of blood in small vessels becomes diameter-dependent as well as hematocrit-dependent (Fahraeus-Lindqvist effect).⁷⁹ This so-called phase separation is also the reason blood cells and plasma tend to distribute unevenly at bifurcations, leading to variations in hematocrit throughout the network. In some cases, this can result in capillaries with low hematocrit or even plasma channels carrying no blood cells at all, which can lead to local areas of hypoxia. If flow regulation is intact, this is normally mitigated via upstream vasodilation⁷⁸; however, in pathophysiological conditions (like sepsis) where blood flow regulation is impaired, areas of hypoxia that persist can lead to organ damage or even failure.⁸⁰

10. Capillary to Mitochondria

The last step in the oxygen transport cascade is at the interface of the capillaries and tissues. This is where oxygen diffuses from the blood to the tissues. Beyond the oxyhemoglobin dissociation curve and a local tissue environment that can facilitate off-loading of oxygen from hemoglobin, this process is largely driven by physical factors. These include, as was the case in the lung, the thickness of membranes, the number of capillaries around a given cell, and the diffusion distance path of oxygen from the capillaries to the mitochondria. The relationship between these factors has been extensively modeled and studied since the time of August Krogh developed the so-called Krogh cylinder model for oxygen transport in skeletal muscle. This remarkable model, while perhaps oversimplified, has shown great explanatory power for 100 years or more as people think about oxygen transport to the tissues.^81,82

As discussed earlier, in the lung a number of things can go wrong at the alveolar pulmonary capillary interface. Likewise things can go wrong at the capillary- tissue interface. Poorly functioning elements of the microcirculation and capillaries can limit vasodilation and delivery of red blood cells to metabolically active tissues. This can be seen in diabetes and inflammatory diseases with notable effects on tissues like the retina, heart, and kidney.

By contrast, as is the case with animals adapted to high altitude the capillary architecture that facilitates the offloading of oxygen at the tissues, endurance exercise training evokes similar adaptations in skeletal muscle. These include angiogenesis throughout the vascular tree but most notably an increase in capillary density around exercise trained skeletal muscle.^83,84 Additionally, there can be a large increase in the mitochondrial content of exercise trained skeletal muscle permitting the mitochondria to operate even more efficiently even at very low PO₂.⁸⁵ Once again some of the adaptive responses to a physiological challenge in humans mimics the evolutionary adaptations made over millions of years in animals primed for life at high altitude. It is important to emphasize that not all adaptations are shared by altitude exposure and exercise training. For example, some native high altitude resistant have structural adaptations in the alveolar capillary interface that improve gas exchange⁸⁶, but this is not observed in response to habitual training.⁸⁷

11. Clinical Applications

At first glance, understanding the physiology of the oxygen cascade in a healthy individual performing exercise appears to have little application to a clinician treating patients. However, we contend that there are considerable similarities and overlap. Exercise and pathology are both stressors to oxygen transport and the regulatory and redundant mechanisms that respond are similar. For example, the most fundamental responses to exercise are increased ventilation and cardiac output, both of which ensure sufficient oxygen is taken up at the lung and transported to the metabolically active tissue. In the case of a disease that hinders the ability to uptake oxygen, the typical response is for ventilation and cardiac output to increase in order to compensate and maintain homeostasis. Specific diseases, some of which have been discussed in detail in our review, may stress one part of the oxygen cascade to a point where the regulatory mechanism is at its sustainable limit. An example would be severe interstitial lung disease where it may not be mechanically or energetically feasible to sustain the increased ventilation required to maintain arterial oxygenation. It is here that pathology actually overlaps with elite athletes. With relevance to the above example, some highly trained athletes have such a metabolic demand for oxygen uptake that they are unable to sustain the required pulmonary ventilation to achieve this. In both these cases there is a mismatch between oxygen demand in capacity. The difference being the athletes encounters this through enhanced demand, whereas the patient is via reduced capacity.

12. Conclusion

The transport of oxygen from the air to the tissues is a multistep process that requires the coordination of many anatomical and physiological elements. Acute and chronic adaptive responses can operate to accommodate the demands of exercise and challenges like environmental hypoxia. During pathophysiological challenges these same responses are repurposed to compensate for diminished function at one or more steps of the oxygen transport cascade. The goal in all cases is to maintain whole body homeostasis by ensuring there is an adequate supply of oxygen at the tissues.

Abbreviations

MET

metabolic equivalent

ODC

oxyhemoglobin dissociation curve

V̇O₂

oxygen uptake