Diffusion, the Alveolar–Capillary Membrane, and Gas Transfer
Ventilation brings gas to the alveoli, and perfusion brings blood to their capillaries. Diffusion then moves gas across the structures separating them. The resulting exchange depends on the pressure gradient, the available pathway, blood uptake, and the time during which gas and blood remain in contact. A low arterial oxygen tension cannot identify which of these steps has failed.
33.1 Gas Crosses a Pathway Rather Than One Isolated Membrane
Oxygen moves from alveolar gas through the surface lining layer, alveolar epithelium, interstitial structures, and capillary endothelium. It then traverses plasma and the red-cell boundary before reaching intracellular hemoglobin. Carbon dioxide moves in the opposite direction, with its transport also influenced by chemical reactions within blood.
The alveolar–capillary membrane is therefore a useful physiological term for a multicomponent pathway. Different components can impose different resistances. A structurally thin tissue barrier does not eliminate the distance between capillary endothelium and red cells or make hemoglobin binding instantaneous.
The barrier is also uneven. Thin exchange regions lie beside thicker regions containing structural tissue and cellular components. Gas transfer through parallel pathways depends strongly on the more conductive paths. An arithmetic average of tissue thickness does not fully describe the barrier's resistance to diffusion.
Weibel and Knight used morphometric measurements and geometric analysis to distinguish thickness related to tissue mass from thickness relevant to gas conductance. Their study showed why the distribution of thickness matters, rather than simply one average microscopic distance. Its experimental measurements were structural observations, not a bedside measurement of a patient's oxygen transfer. 1
The area available in life is also functional. An alveolar surface contributes effectively only when gas reaches it and blood and red cells are appropriately exposed to it. Loss of aeration, vascular obstruction, and uneven red-cell distribution can reduce useful exchange without producing an identical change in the anatomical surface seen on a tissue section.
Separating structure from function prevents a misleading interpretation of imaging. Thickened interstitium can plausibly hinder transfer, but its contribution to hypoxemia must be assessed alongside aeration, ventilation–perfusion matching, blood uptake, and circulation.
33.2 Fick's Diffusion Relationship Defines the Driving Forces
For a simplified barrier, gas transfer increases when the available area or pressure gradient increases. Transfer decreases when the diffusion distance increases. The gas's solubility and molecular diffusivity determine how readily it moves through that material.
Gas permeability here combines the material's solubility for the gas with its diffusion coefficient. The relationship assumes a defined pathway with suitably stable conditions. It describes how a homogeneous barrier behaves before adding the full complexity of capillary blood and regional lung structure.
Diffusing capacity expresses transfer per unit pressure difference. For a specified gas and physiological model:
Traditional diffusing-capacity units are mL gas/min/mm Hg. Gas-transfer rate is then in mL/min, with the gas-volume convention stated. Laboratory reporting commonly uses standard temperature and pressure, dry, abbreviated STPD. SI reporting uses molar quantities and pressure in kPa, so units must be converted before comparing values.
Consider an illustrative oxygen conductance of 20 mL O₂/min/mm Hg and an effective mean oxygen gradient of 10 mm Hg. The conductance states how much oxygen crosses each minute for each mm Hg of driving difference. Multiplying gives 20 × 10 = 200 mL O₂/min. This is a teaching calculation, not a measured resting oxygen consumption or a DLCO result.
The effective mean gradient is not simply alveolar oxygen minus final arterial oxygen. Capillary oxygen tension rises along the pathway, so the local gradient becomes smaller as blood is oxygenated. Different regions also have different gradients. An appropriate average requires a model of that changing exchange.
In the membrane-only model, doubling thickness at fixed area and material properties halves conductance. In vivo, blood resistance, recruitment, and the distribution of pathways also contribute. The same proportional statement cannot automatically be applied to total measured lung transfer.
33.3 Diffusion Limitation Means Equilibration Remains Incomplete
As venous blood enters a ventilated alveolar capillary, its oxygen tension is usually below alveolar oxygen tension. Oxygen diffuses into blood. Dissolved oxygen tension rises while hemoglobin binds oxygen and increases total content. The difference between alveolar and capillary tension drives continuing transfer.
If capillary oxygen tension approaches alveolar tension before blood leaves the exchange region, the local driving gradient becomes very small. Additional oxygen uptake then mainly requires additional blood flow or a change in gas and blood conditions. This behavior is called perfusion-limited exchange.
Diffusion limitation occurs when blood leaves without adequate equilibration with the relevant alveolar gas. Increasing conductance or available contact time could then increase uptake. A persistent end-capillary gradient is the defining feature, rather than the mere presence of a thickened membrane or a low arterial oxygen tension.
The labels depend on the gas and circumstances. Carbon monoxide binds strongly to hemoglobin, keeping its dissolved capillary tension low during a standard low-concentration transfer test. Uptake can continue despite substantial accumulated blood content. This makes it useful for investigating transfer conductance, but it does not mean oxygen behaves identically under the same conditions.
For oxygen, a healthy resting lung usually has enough reserve to approach local equilibration. Disease, exercise, and reduced inspired oxygen can challenge that reserve. The same person can therefore show little diffusion limitation at rest yet develop disequilibrium under another condition.
Torre-Bueno and colleagues compared measured oxygenation with that predicted from inert-gas-derived ventilation–perfusion inequality during exercise at sea level and simulated altitude. Their findings supported oxygen diffusion disequilibrium during exercise at higher simulated altitudes, rather than a fixed limitation under every condition. The distinction was inferred through physiological measurements and modeling, not direct sampling of every end-capillary region. 2
33.4 Transit Time Must Be Interpreted With Recruitment and Distribution
Capillary transit time describes how long blood remains within an exchanging capillary pathway. A simple mean can be estimated by dividing the volume of blood in the relevant capillary bed by its blood flow. Volume and flow must refer to the same compartment.
Take an illustrative capillary blood volume of 80 mL and flow of 6 L/min. Convert flow first: 6,000 mL/min divided by 60 seconds/min equals 100 mL/s. The mean transit time is 80/100 = 0.8 seconds. If flow doubles to 12 L/min, or 200 mL/s, while the assumed volume stays at 80 mL, the mean becomes 80/200 = 0.4 seconds.
The calculation explains why increasing flow can shorten contact time. It does not establish the actual transit of every red cell. The capillary network contains pathways with different lengths, flows, red-cell distributions, and recruitment states. A satisfactory average can coexist with shorter transit in some pathways.
Exercise also recruits transfer reserve. Capillary distension, recruitment, and changes in red-cell exposure can increase effective conductance and blood volume. Capillary blood volume cannot be assumed to remain fixed while flow rises in a living lung.
Hsia and colleagues modeled red-cell distribution within pulmonary capillaries and showed that rearranging cells could alter gas conductance even at a fixed overall red-cell number. Their analysis supports a role for microvascular distribution in recruitment of transfer capacity. It does not directly measure the contribution of that mechanism in an individual exercising patient. 3
A diseased lung may have less reserve because functioning surface or capillary bed has been lost. Shorter transit, lower alveolar oxygen, and reduced conductance can then combine to impair equilibration. Desaturation with exertion is compatible with this mechanism, but ventilation–perfusion inequality, vascular disease, and venous oxygen changes can also contribute.
33.5 Edema and Fibrosis Alter Several Exchange Mechanisms
Interstitial fluid can increase diffusion distance and change the mechanical environment of small airways and capillaries. Alveolar flooding can remove a region from effective ventilation while perfusion continues. Compression and altered vascular pressures can redistribute blood. Edema can therefore impair exchange through diffusion, low ventilation–perfusion ratios, shunt, and circulatory effects together.
Fibrosis can thicken and distort the exchange pathway, reduce the available alveolar surface, and alter the vascular bed. Reduced lung expansion and uneven mechanics can also change ventilation distribution. A low transfer test or exertional hypoxemia may reflect several of these abnormalities.
Agustí and colleagues studied patients with idiopathic pulmonary fibrosis using inert gases and respiratory measurements at rest and during exercise. They identified contributions from both ventilation–perfusion inequality and oxygen diffusion limitation, with a larger diffusion component during exercise. The study illustrates coexistence rather than supporting a membrane-only explanation for fibrotic hypoxemia. 4
In acute lung injury, a radiographic infiltrate does not specify the mechanism of gas-exchange failure. Dantzker and colleagues measured ventilation–perfusion distributions in ARDS and found substantial perfusion of shunt or extremely low-ratio regions. Their work demonstrates that severe hypoxemia in an injured, edematous lung can arise prominently from regional ventilation failure. 5
The distinction has practical consequences. Increasing inspired oxygen can strengthen the diffusion gradient in ventilated units. It cannot directly ventilate a flooded or closed region. A pressure change that improves aeration may also alter perfusion and cardiac output. The improvement therefore cannot be assigned automatically to a thinner diffusion barrier.
Describe the affected processes separately: altered barrier conductance, loss of ventilated area, vascular abnormalities, and flow through poorly ventilated regions. Their relative contributions can change during illness and recovery.
33.6 DLCO Measures Carbon-Monoxide Uptake Under Test Conditions
The diffusing capacity of the lung for carbon monoxide, DLCO, is also called the carbon-monoxide transfer factor. It measures uptake of a small test amount of carbon monoxide relative to its driving pressure. It is a functional gas-transfer measurement rather than a direct microscopic measurement of barrier thickness.
In the single-breath approach, the subject inhales test gas containing carbon monoxide and an inert tracer, holds the breath for a defined interval, and exhales for analysis. Carbon-monoxide disappearance estimates uptake. Tracer dilution helps estimate the accessible alveolar volume and initial gas concentration.
Forster and colleagues investigated carbon-monoxide absorption during breath-holding and developed the physiological basis of this measurement. Their experiments examined disappearance from alveolar gas after inspiration, linking the observed uptake to an estimated conductance. 6
A standard calculation assumes that dissolved capillary carbon-monoxide pressure is sufficiently small relative to alveolar pressure. Background carboxyhemoglobin and carbon-monoxide back pressure can modify that assumption. The concentration measurement also depends on appropriate gas mixing and sampling.
The maneuver samples the lung reached by the inspired test gas. Incomplete inspiration, uneven ventilation, leaks, and unsuitable breath-hold or collection conditions can change the result. Changes in lung volume or intrathoracic pressure during the maneuver can also affect capillary conditions.
The ERS/ATS technical standard specifies equipment, maneuver, calculation, and quality requirements to make measurements interpretable. Its central implication is that a numerical DLCO must be considered with test quality and measurement conditions, not simply compared with a remembered normal value. 7
This laboratory maneuver is not a routine direct measurement of oxygen diffusion during unstable ICU ventilation. A previous low DLCO can inform the patient's pulmonary history, but it cannot independently identify the cause of a new oxygenation problem or quantify the present oxygen conductance.
33.7 Hemoglobin and Accessible Volume Are Part of the Result
Hemoglobin provides the major sink for carbon monoxide entering capillary blood. A lower hemoglobin concentration can reduce measured uptake even when the tissue barrier has not changed. The result must therefore be interpreted with the patient's hemoglobin and the laboratory's adjustment method.
Cotes and colleagues examined transfer factor and exercise responses in iron-deficiency anemia before and after iron or placebo treatment. Their study placed blood composition within the physiological assessment of transfer. A low value in an anemic patient should not be assigned automatically to otherwise unproven interstitial disease. 8
Accessible alveolar volume, V_A, adds another dimension. The transfer coefficient, KCO, expresses carbon-monoxide uptake per unit accessible volume in the test. Their relationship is:
If KCO is expressed in mL CO/min/mm Hg/L and V_A in L, their product is in mL CO/min/mm Hg. An illustrative KCO of 5 and accessible volume of 4 L give a DLCO of 5 × 4 = 20. A KCO of 4 with volume of 5 L also gives 20. The same total result can therefore arise from different component values.
Dividing by volume does not simply correct DLCO for a small lung. In healthy subjects, Stam and colleagues found that reducing alveolar volume lowered total DLCO while increasing DLCO per unit volume. The relationship is not a constant proportional scaling. A preserved KCO cannot automatically exclude disease when V_A is low. 9
V_A measured by tracer dilution is also not necessarily identical to total lung capacity measured by another method. Poorly communicating regions may contribute little to the sampled volume. Interpret V_A, KCO, lung volumes, spirometry, hemoglobin, and imaging together.
33.8 Roughton–Forster Partitioning Treats Resistances in Series
Gas must cross the barrier and then be taken up by capillary blood. The Roughton–Forster model represents these as resistances in series. A resistance is the reciprocal of conductance, so membrane and blood resistances add even though their conductances do not.
D_MCO is the membrane component of conductance for carbon monoxide. V_c is the participating pulmonary capillary blood volume. θ_CO, pronounced theta, represents carbon-monoxide uptake conductance per unit blood volume under the assumed blood conditions. Multiplying theta by capillary volume gives the blood component of conductance in the same units as DLCO and D_MCO.
Roughton and Forster developed this framework to distinguish membrane transfer from finite blood uptake. Their work showed why chemical combination and red-cell processes must be considered alongside passage through the tissue barrier. 10
Consider illustrative membrane conductance of 60 and blood conductance of 60 mL CO/min/mm Hg. Their resistances are each 1/60. Adding gives 1/60 + 1/60 = 1/30, so total DLCO is 30. It is smaller than either component because gas encounters both resistances in sequence.
Now halve membrane conductance to 30 while blood conductance stays 60. The sum becomes 1/30 + 1/60 = 1/20. Total DLCO becomes 20, not 15. Halving one component conductance does not halve the total when another resistance remains in series.
Partitioning requires assumptions about theta, hemoglobin, oxygen tension, participating blood volume, and the model's representation of heterogeneous pathways. Methods using different inspired oxygen conditions rely on how oxygen changes carbon-monoxide uptake by blood. The derived membrane and volume terms are estimates within that framework, not independent tissue and blood-volume images.
33.9 Nitric-Oxide Transfer Adds Information but Retains Blood-Side Assumptions
Nitric oxide is taken up rapidly by blood and can be measured as a separate test-gas transfer capacity, DLNO. Because its uptake kinetics differ from carbon monoxide, comparing the two gases can help investigate the balance between membrane and blood components.
Guenard and colleagues measured nitric-oxide and carbon-monoxide transfer in healthy subjects and used their different uptake behavior to estimate membrane conductance and capillary blood volume. Their approach treated the blood resistance for nitric oxide as very small, allowing the nitric-oxide measurement to approximate a membrane-related term. 11
Very rapid uptake does not mean that blood presents literally no resistance. Nitric oxide still encounters plasma, the red-cell boundary, intracellular transport, and hemoglobin reactions. The resulting value can depend on red-cell properties and on the kinetic coefficients chosen for analysis.
Chakraborty and colleagues modeled diffusion and reaction during uptake of oxygen, carbon monoxide, carbon dioxide, and nitric oxide by red cells. Their analysis showed how cell geometry, surrounding plasma, diffusion, and reaction can jointly determine blood-side conductance. This provides a reason to state kinetic assumptions rather than treating a test gas as a pure membrane probe. 12
DLNO and DLCO also require appropriate gas delivery, sampling, lung volume, and quality control. Their ratio can add physiological information, but it does not uniquely separate every vascular disorder from every interstitial disorder. Disease can alter several components simultaneously.
These measurements are particularly useful when interpreted as complementary functional observations. They do not independently establish how much of an ICU patient's oxygenation failure comes from disequilibrium rather than mismatch or shunt.
33.10 End-Capillary Equilibrium Does Not Guarantee Normal Arterial Oxygen
Local equilibrium compares capillary blood with the alveolar gas beside it. It does not imply that the alveolar gas has a normal oxygen tension. A low ventilation–perfusion unit can fully equilibrate its blood with a low alveolar oxygen tension and still deliver poorly oxygenated blood to the pulmonary veins.
An idealized illustrative unit might have alveolar oxygen tension of 55 mm Hg. If its end-capillary blood also reaches 55 mm Hg, there is no remaining local oxygen gradient. Its exchange is not diffusion-limited by that criterion, yet its outflow may contribute to arterial hypoxemia.
A shunt pathway creates another distinction. Blood passing through a nonventilated region does not encounter an adequate fresh-gas source. Calling the resulting deficit diffusion failure confuses lack of gas access with failure to equilibrate across an available exchanging pathway.
After regional blood streams merge, arterial oxygen content reflects their flow-weighted mixture. The final arterial tension is then related to that content through hemoglobin binding and other blood conditions. It is not a direct measurement of the terminal pressure in one representative alveolar capillary.
Research can compare measured oxygenation with values predicted from measured ventilation–perfusion distributions. A residual discrepancy can support diffusion disequilibrium under the model's assumptions, as in the exercise and fibrosis studies discussed earlier. It does not make the same separation possible from a single ordinary arterial sample.
Supplemental oxygen can strengthen the diffusion gradient in ventilated units, but it can also improve low-ratio exchange. Thus, an oxygen response does not uniquely diagnose diffusion limitation. Exercise desaturation and a widened alveolar–arterial difference likewise require interpretation rather than a single mechanistic label.
33.11 Carbon Dioxide Crosses Readily but Still Requires Effective Exchange
Carbon dioxide is more soluble in the exchange pathway than oxygen, giving it a greater capacity for transfer across tissue for a comparable partial-pressure gradient. Its molecular size alone does not determine this property. Solubility and diffusivity both enter the permeability relationship.
This helps explain why substantial oxygen-transfer impairment can occur while carbon dioxide remains normal or low. Increased ventilation can also compensate for carbon-dioxide elimination in accessible, adequately perfused regions. A normal PaCO₂ therefore does not prove that the gas-transfer pathway is entirely normal.
Greater diffusibility is not immunity from exchange failure. Carbon dioxide must be released from blood, cross the pathway, and leave through ventilation. Blood reactions involving bicarbonate, carbonic anhydrase, and hemoglobin contribute to the kinetics. The red-cell analysis discussed above treats carbon-dioxide transfer as a coupled transport-and-reaction process rather than an unrestricted passage through tissue.
In clinical lung injury, carbon-dioxide retention often reflects inadequate effective alveolar ventilation relative to production, including the effects of dead space, mismatch, and mechanical limits. It should not automatically be assigned to a thicker membrane. Conversely, low PaCO₂ can require substantial compensatory ventilation and does not establish that the respiratory system is operating efficiently.
Consider an edematous patient with hypoxemia and rising PaCO₂. Review ventilation delivery, dead-space changes, respiratory effort, airway obstruction, and carbon-dioxide production alongside the structural lung injury. The presence of edema is relevant, but it does not remove the need to identify these other constraints.
Chapter 35 develops carbon-dioxide carriage and elimination in greater detail. Here, its importance is to prevent the oxygen and carbon-dioxide results from being treated as interchangeable measures of membrane thickness.
33.12 Build the Bedside Explanation From Several Measurements
Consider an illustrative patient with established fibrotic lung disease, a previously reduced DLCO, and new severe hypoxemia during pneumonia. The earlier transfer test identifies limited gas-transfer reserve under its measurement conditions. It does not establish the mechanism of the new deterioration.
First verify inspired oxygen and ventilation. Review regional aeration, airway patency, imaging, and the clinical course. New flooded or poorly ventilated regions may introduce shunt or low ventilation–perfusion ratios on top of the chronic structural disease.
Next assess the blood and circulation. Anemia can reduce oxygen content and alter a transfer measurement. Reduced cardiac output can lower incoming venous oxygen when tissue extraction increases. Pulmonary vascular disease can alter available capillary bed and distribution. These processes can amplify hypoxemia without being equivalent to an isolated increase in barrier thickness.
If a laboratory transfer result is available, examine its quality, hemoglobin context, accessible volume, and transfer coefficient. Compare like conditions across time. A change in inspiratory volume or sampling quality can alter a result even when the underlying disease has not changed to the same degree.
Exercise-related deterioration or improvement with oxygen can support a physiological hypothesis, but neither isolates diffusion. Follow the combined response of oxygenation, carbon-dioxide elimination, mechanics, and perfusion. Improved arterial tension accompanied by impaired cardiac output may not improve systemic oxygen delivery.
The most defensible explanation may include several mechanisms: reduced structural conductance, diminished transfer reserve, uneven gas and blood delivery, and a changed venous input. Diffusion remains an essential part of that explanation when it is connected to the available pathway, local gradient, blood uptake, and contact time.
References
1. Weibel ER, Knight BW. A morphometric study on the thickness of the pulmonary air-blood barrier. J Cell Biol. 1964;21(3):367–384. doi:10.1083/jcb.21.3.367
2. Torre-Bueno JR, Wagner PD, Saltzman HA, Gale GE, Moon RE. Diffusion limitation in normal humans during exercise at sea level and simulated altitude. J Appl Physiol. 1985;58(3):989–995. doi:10.1152/jappl.1985.58.3.989
3. Hsia CCW, Johnson RL Jr, Shah D. Red cell distribution and the recruitment of pulmonary diffusing capacity. J Appl Physiol. 1999;86(5):1460–1467. doi:10.1152/jappl.1999.86.5.1460
4. Agustí AG, Roca J, Gea J, Wagner PD, Xaubet A, Rodriguez-Roisin R. Mechanisms of gas-exchange impairment in idiopathic pulmonary fibrosis. Am Rev Respir Dis. 1991;143(2):219–225. doi:10.1164/ajrccm/143.2.219
5. Dantzker DR, Brook CJ, Dehart P, Lynch JP, Weg JG. Ventilation-perfusion distributions in the adult respiratory distress syndrome. Am Rev Respir Dis. 1979;120(5):1039–1052. doi:10.1164/arrd.1979.120.5.1039
6. Forster RE, Fowler WS, Bates DV, Van Lingen B. The absorption of carbon monoxide by the lungs during breath-holding. J Clin Invest. 1954;33(8):1135–1145. doi:10.1172/JCI102987
7. Graham BL, Brusasco V, Burgos F, et al. 2017 ERS/ATS standards for single-breath carbon monoxide uptake in the lung. Eur Respir J. 2017;49(1):1600016. doi:10.1183/13993003.00016-2016
8. Cotes JE, Dabbs JM, Elwood PC, Hall AM, McDonald A, Saunders MJ. Iron-deficiency anaemia: its effect on transfer factor for the lung (diffusing capacity) and ventilation and cardiac frequency during sub-maximal exercise. Clin Sci. 1972;42(3):325–335. doi:10.1042/cs0420325
9. Stam H, Hrachovina V, Stijnen T, Versprille A. Diffusing capacity dependent on lung volume and age in normal subjects. J Appl Physiol. 1994;76(6):2356–2363. doi:10.1152/jappl.1994.76.6.2356
10. Roughton FJW, Forster RE. Relative importance of diffusion and chemical reaction rates in determining rate of exchange of gases in the human lung, with special reference to true diffusing capacity of pulmonary membrane and volume of blood in the lung capillaries. J Appl Physiol. 1957;11(2):290–302. doi:10.1152/jappl.1957.11.2.290
11. Guenard H, Varene N, Vaida P. Determination of lung capillary blood volume and membrane diffusing capacity in man by the measurements of NO and CO transfer. Respir Physiol. 1987;70(1):113–120. doi:10.1016/S0034-5687(87)80036-1
12. Chakraborty S, Balakotaiah V, Bidani A. Diffusing capacity reexamined: relative roles of diffusion and chemical reaction in red cell uptake of O₂, CO, CO₂, and NO. J Appl Physiol. 2004;97(6):2284–2302. doi:10.1152/japplphysiol.00469.2004