Exercise: The Whole Body at Work

At rest, a healthy young man of about 70 kg takes up about 250 mL of oxygen per minute. At the end of a maximal run he may take up ten or more times as much. That oxygen passes through lungs, heart, blood and capillaries before a mitochondrion uses it. Which link sets the ceiling?

This lesson splits that rise into its factors, follows each organ system, and weighs the human experiments on the ceiling.

VO2max is a ceiling, and Fick splits it in two

Oxygen uptake (VO2) is the oxygen the whole body takes up per minute. In an incremental test the work rate on a treadmill or cycle ergometer rises in steps, and expired-gas analysis of air breathed through a mouthpiece gives VO2 as oxygen breathed in minus oxygen breathed out. VO2 climbs with the work rate until further increases in speed or work rate no longer raise it. Hill and Lupton (Quarterly Journal of Medicine, 1923) described that ceiling in runners as the maximal oxygen uptake (VO2max); without a clear plateau, laboratories report peak VO2.

Breath-by-breath values are averaged before the maximum is read, and a 2020 review found that the averaging strategy alone can shift the reported maximum by 4 to 10 percent. Keep absolute VO2 (L per min) apart from mass-specific VO2 (mL per kg of body mass per min), which compares people of different sizes.

Cardiac output and venous return used Fick’s principle to measure flow. Here it splits oxygen uptake:

VO2 = cardiac output x (CaO2 - CvO2), and cardiac output = heart rate x stroke volume

CaO2 is the oxygen content of arterial blood and CvO2 that of mixed-venous blood, pulmonary-artery blood in which returns from every organ have mixed, in mL of oxygen per L of blood. These are contents, not partial pressures: bound plus dissolved oxygen, read from the curve in oxygen transport. Femoral venous blood from one leg is not mixed-venous. CaO2 - CvO2 is the oxygen extracted per litre; divided by CaO2 it is the extraction fraction.

Worked problem: splitting the rise with Fick

Numbers are illustrative, except the resting uptake of about 250 mL per min for a healthy 70-kg young man (Joyner and Casey, 2015). Assume steady state and no shunt.

1. Rest. Cardiac output 5.0 L per min, CaO2 200 mL per L, CvO2 150 mL per L. The arteriovenous difference is 200 - 150 = 50 mL per L, so VO2 = 5.0 L per min x 50 mL per L = 250 mL per min.
2. Maximal exercise, cardiac output. Heart rate 195 beats per min and stroke volume 100 mL per beat give 195 x 100 = 19,500 mL per min, or 19.5 L per min.
3. Maximal exercise, extraction. CaO2 200 mL per L and CvO2 40 mL per L give a difference of 160 mL per L, so VO2max = 19.5 x 160 = 3,120 mL per min, 3.12 L per min.
4. Per kilogram. 3,120 / 70 = 44.6 mL per kg per min.
5. The split. VO2 rose 3,120 / 250 = 12.5-fold: a 3.9-fold rise in cardiac output (19.5 / 5.0) times a 3.2-fold rise in extraction (160 / 50). Check: 3.9 x 3.2 = 12.48.
6. Check CaO2. Hemoglobin 150 g per L, 97 percent saturated, binding 1.34 mL of oxygen per gram (a common value; published values run from about 1.34 to 1.39): 1.34 x 150 x 0.97 = 195 mL per L. Dissolved oxygen at a PO2 of 100 mmHg adds 0.03 mL per L per mmHg x 100 = 3 mL per L. Total 198 mL per L, close to the 200 used.

Heart rate 195 beats per min Stroke volume 100 mL per beat Hemoglobin 15.4 g per dL (154 g per L) Fraction of arterial oxygen extracted 80%

Lesson’s resting example Lesson’s maximal example

Cardiac output

19.5 L per min

Arterial content, CaO₂

200 mL per L

Mixed-venous content, CvO₂

40 mL per L

Oxygen extracted per litre

160 mL per L

VO₂

3.12 L per min

VO₂ for 70 kg

44.6 mL per kg per min

In this sketch: steady state, no shunt, a 70-kg person. Arterial content = 1.34 × hemoglobin × 10 × 0.97 mL per L (1.34 mL per g is a common value, and published values run from about 1.34 to 1.39; 97 percent saturated; dissolved oxygen is ignored, which the lesson puts at about 3 mL per L at a PO₂ of 100 mmHg). Mixed-venous content = arterial content × (1 − extraction). Hemoglobin 15.4 g per dL is chosen in this sketch so that CaO₂ is about 200 mL per L, the lesson’s value; the lesson’s own check uses 150 g per L, which gives 195 mL per L. The resting marker, about 250 mL per min, is the lesson’s value for a healthy 70-kg young man; the lesson’s other numbers are illustrative. The lesson gives a resting cardiac output of 5.0 L per min but not its split, so the resting button uses 60 beats per min × 83 mL, made up for this sketch. Slider ranges are made up too, and the sketch does not say which combinations people reach.

VO₂ = heart rate × stroke volume × CaO₂ × extraction fraction. A product has the same value for many mixtures of factors, and it falls if any one factor is lowered while the others are held.

The heart pumps more, and the flow goes to working muscle

Cardiac output rises through heart rate and stroke volume. Stroke volume rises with preload, the end-diastolic stretch of the ventricle, and contractility, the strength of contraction at a given stretch (pressure-volume loops). Sympathetic outflow rises (hypothalamus and autonomic control), and the baroreflex operates around a higher pressure (regulating arterial pressure).

Where the flow goes matters as much. In heavy exercise, sympathetic vasoconstriction can reduce renal and splanchnic blood flow. The same nerves reach active muscle, but contraction can blunt their constricting effect there, which is called functional sympatholysis. Active muscle also dilates through local signals (exercise hyperemia), and a 2015 review concluded that no single vasodilator explains it (resistance, compliance and local flow). Picture a water system at peak demand: the pumps run much harder while the district on fire opens its hydrants, and closing a few suburban valves helps hold the pressure. Most of the extra flow to muscle comes from the larger cardiac output, not from the gut and kidneys.

How much flow could muscle take if the heart were not the limit? Per Andersen and Bengt Saltin (1985, Journal of Physiology 366:233) had five healthy subjects do dynamic knee extension with one leg at 10 to 60 W, so the active muscle was essentially one quadriceps, and measured femoral venous flow by cold-saline thermodilution. If muscle vascular capacity set the whole-body ceiling, a small muscle alone should reach no more flow per kilogram than in whole-body exercise. If the heart set it, a muscle too small to strain the heart should take far more. Femoral flow rose with workload to an average of 5.7 L per min, and the estimated maximal perfusion was about 2.5 L per kg of active muscle per min (per kilogram of quadriceps, not of body mass).

That showed a large reserve of muscle vascular capacity and argued against muscle dilator capacity as the ceiling in these subjects. It did not test whole-body exercise or the dilator mechanism, and the per-kilogram figure depends on the estimated quadriceps mass.

Worked problem, part 2: the argument in numbers

Except for the 2.5 L per kg per min, values are made up. Suppose running uses 20 kg of muscle and 80 percent of part 1’s 19.5 L per min maximal cardiac output reaches it. At 2.5 L per kg per min the muscles could accept 2.5 x 20 = 50 L per min; the heart sends 0.80 x 19.5 = 15.6 L per min. The ratio is 50 / 15.6 = 3.2 (check: 15.6 x 3.2 = 49.9). Muscle could accept three times what the heart delivers, so its vessels must be partly constricted, and here the heart sets the ceiling. This assumes the one-leg maximum applies to every working muscle, the active mass is known and arterial pressure matches that of the one-leg test, where it rose from about 100 to 130 mmHg at the higher workloads.

Breathing keeps arterial gases steady, until it does not

Ventilation rises at once at exercise onset. At low to moderate intensity it keeps pace with carbon dioxide output, and arterial PO2 and PCO2 stay near their resting values of about 100 and 40 mmHg at sea level. In heavy exercise ventilation can outpace carbon dioxide output (hyperventilation), and arterial PCO2 falls. The control is in control of breathing.

The healthy lung usually keeps arterial oxygen up even at maximal work. Jerome Dempsey, P. G. Hanson and K. S. Henderson (1984) sampled arterial blood in 16 highly trained runners during treadmill exercise at sea level. In eight of them, arterial PO2 fell during exercise by 21 to 35 mmHg. The worst falls went with a wide alveolar-arterial oxygen difference and, in some runners, too little hyperventilation. The authors judged uneven ventilation-perfusion and shunt unlikely and proposed, by exclusion, a diffusion limit from very short red-cell transit times in the lung capillaries; the cause is still debated (ventilation and perfusion explains the A-a difference). It is a finding in some highly trained runners, not in healthy people generally.

Whether the hard-working breathing muscles, fed from the same cardiac output, take flow from the legs is the design problem.

Fuel shifts toward carbohydrate, and lactate is a fuel

Which fuel muscle burns depends on intensity (energy systems are in skeletal muscle mechanics). Romijn and colleagues (1993) studied five trained subjects at 25, 65 and 85 percent of VO2max with stable-isotope tracers and indirect calorimetry. Plasma glucose uptake and muscle glycogen oxidation rose with intensity, and total fat oxidation was highest at the middle intensity.

During prolonged exercise plasma insulin falls and glucagon rises, favouring glucose release from the liver (insulin, glucagon and blood glucose), while contracting muscle takes up more glucose by a route that does not require insulin (integration of metabolism).

Lactate is a fuel, not waste. Some glycolytic pyruvate becomes lactate, which leaves the fibres that make it and is oxidized by other fibres and the heart or made into glucose by the liver. The lactate threshold is the workload above which blood lactate accumulation speeds up as work rises. It is not where oxygen runs out, and it has no universal value (not a fixed 4 mmol per L, percentage of VO2max or ventilatory threshold). Messonnier and colleagues (2013) used tracers in six trained and six untrained men, and a lactate clamp, an infusion of lactate that holds its blood concentration, in the trained men only. At each group’s own threshold, lactate clearance was 34 percent higher in the trained men, and clearance reached its peak below the threshold. In a study of six trained and six untrained men, Emhoff and colleagues (2013) estimated with tracers that blood lactate supplied about one-third of carbohydrate-derived energy during 60 minutes of exercise at the lactate threshold. Both studies compared trained with untrained men, so the differences are associations.

Early in exercise, heat production can outrun heat loss and core temperature rises. Skin blood flow carries heat from the core to the skin, and sweat cools only when it evaporates (skin, sweat and human thermoregulation). Heat also strains the heart: in 15 endurance-trained cyclists, González-Alonso and colleagues (1997) found that hyperthermia alone or about 4 percent dehydration alone lowered stroke volume by 7 to 8 percent, and the two together lowered cardiac output by 13 percent.

What limits VO2max, and what training changes

In healthy people doing whole-body exercise at sea level, the evidence points to oxygen delivery, mainly cardiac output, rather than muscle oxygen use. Mortensen and colleagues (2008) found that in 13 trained men cycling to exhaustion at a supramaximal work rate, systemic and leg flow peaked after about 60 to 90 seconds and then levelled off, which the authors read as cardiac limits plus muscle vasoconstriction.

The modern test came from training. Montero, Lundby and colleagues (2015) trained 16 previously untrained healthy men for six weeks, measuring peak cardiac output, blood volumes and muscle biopsies. If the gain came from muscle, removing the extra red cells afterwards should leave most of it; if from delivery, removal should erase it. Peak VO2 rose 9 plus or minus 8 percent (mean and SD), mitochondrial volume density 43 plus or minus 30 percent and capillary-to-fibre ratio 18 plus or minus 16 percent. The authors report that returning red-cell volume to baseline by phlebotomy returned peak VO2 to its pretraining level; a later study of 12 untrained adults, seven of them men, after 10 weeks of training (Skattebo and colleagues, 2020) found the gain preserved after phlebotomy, so the result is not settled. It does not show that mitochondria never limit VO2, or anything about women, older adults or athletes; and phlebotomy lowers blood volume and preload too, so it cannot separate oxygen-carrying capacity from volume (blood cells and hemostasis covers red-cell mass).

Strength training differs. Jenkins and colleagues (2017) randomized 26 young men to six weeks of leg extension at 80 or 30 percent of their one-repetition maximum. Ultrasound muscle thickness rose similarly in both groups, but strength and voluntary activation, how fully the nervous system drives the muscle in a maximal effort (motor units and reflexes), were greater with the heavy load, which suggests a neural contribution. In rats, the MoTrPAC consortium (2024) mapped molecular training responses in blood and 18 solid tissues, with sex differences in many: associations, not causes. Montero and Lundby (2017) found that two extra weekly sessions removed non-response as they defined it (a change in maximal power within the measurement error of plus or minus 3.96 percent).

Heart failure with reduced ejection fraction tests the Fick split. Against a healthy person of the same age, peak stroke volume is low and rises little (weak contractility) and peak heart rate is often lower, so peak cardiac output is low. At a given submaximal work rate flow is low for the demand, so extraction is higher and mixed-venous oxygen content falls earlier. Low peak cardiac output is a main reason peak VO2 is low, but oxygen transport from blood into muscle is also reduced (Esposito and colleagues, 2010). Higher sympathetic outflow constricts muscle vessels more, and lactate accumulates at a lower absolute work rate. Chronic heart failure also changes muscle (lower mitochondrial volume and enzymes, in some studies fewer capillaries per fibre, and a shift in fibre type), which contributes a peripheral limit, and ventilation per litre of carbon dioxide output is higher.

Design problem

At maximal cycling a trained athlete’s breathing muscles work hard. Does that work take blood flow from the legs? Design an experiment in trained male cyclists and say what each result would mean.

One way to do it

Hold the cycling power constant at or near maximum and change the work of breathing: on separate visits, in random order, unload inspiration with a ventilator that assists each breath, add an inspiratory resistance, or use a sham setting.

Measure esophageal pressure (to quantify the work of breathing and the swing in chest pressure), femoral venous flow by thermodilution, cardiac output and arterial pressure (leg vascular conductance = flow / pressure), leg and whole-body VO2, arterial blood gases (to exclude a change in arterial gases), and leg noradrenaline spillover as a marker of sympathetic outflow.

If breathing competes for flow, unloading should raise leg flow and conductance and loading should lower them. Measure cardiac output too, because changing the work of breathing can change total cardiac output as well as where it goes. If cardiac output rose to supply both, leg flow would not change.

Harms and colleagues in Dempsey’s laboratory (Journal of Applied Physiology 82:1573, 1997) ran a version of this in seven trained male cyclists, measuring leg flow but not cardiac output. Reducing inspiratory work during maximal cycling increased leg blood flow, and a change of about 50 percent in respiratory work went with a change of about 2 L per min in leg flow in the opposite direction. A 1998 follow-up in eight cyclists measured cardiac output and found it about 9 percent lower with unloading, so unloading did not simply move a fixed cardiac output to the legs; the authors estimated that up to 14 to 16 percent of cardiac output goes to the breathing muscles at maximal exercise. In Harms and colleagues’ data, leg vascular resistance tracked norepinephrine spillover, which fits a proposed reflex from the working respiratory muscles that raises sympathetic outflow to the legs. Seven trained men at maximal exercise say nothing about moderate exercise or women.