Of all the tests performed by pulmonary function technologists, single-breath diffusion capacity (Dlco) is the most complex and requires the most attention and skill for its successful completion. To complete accurate Dlco studies, many factors must be considered.

Physiology of Diffusion

Dlco is a direct measurement, by indirect means, of the integrity of the alveolar capillary membrane. The lung contains 5.5 million or so alveoli that take part in gas exchange. Each alveolus has a polygonal shape and consists of squamous cells called type I pneumocytes. Wrapped loosely around each alveolus (or group of alveoli descending from a single respiratory bronchiole) is a fine net of capillaries bringing blood directly from the parallel branch of the pulmonary artery.

Inhaled oxygen in the alveolar spaces diffuses or transfers according to the law of mass action across both the inner and outer membranes of the alveoli, through the thin interstitial space, through the inner and outer membranes of the capillaries, into the blood plasma, and then through the erythrocyte membrane, where it is finally bonded chemically to the hemoglobin (Hb) molecule. Carbon dioxide takes the opposite course from the blood into the alveoli, where it is exhaled.

This diffusion into and out of the alveoli takes place rapidly, and its efficiency depends upon the integrity of the alveolar capillary membrane, as well as upon the rate of blood flow through the capillaries, in addition to the concentration of Hb in the blood and the barometric pressure (Pb). Diseases that reduce Dlco include lung resection, emphysema, interstitial lung disease, pulmonary edema, pulmonary vasculitis, and hypertension. Diseases that increase Dlco include polycythemia, left-to-left shunting, pulmonary hemorrhage, and asthma.

Carbon monoxide binds with Hb to form carboxyhemoglobin (COHb) 70 times more rapidly than oxygen binds with Hb to form oxyhemoglobin. The half-life of COHb is 8 hours. Because of the high binding capacity of carbon monoxide to Hb, this gas is used in the Dlco test to ascertain the condition of the alveolar capillary membrane. The test gas used consists of 0.3% carbon monoxide, either 10% helium or 0.3% methane, 21% oxygen, and nitrogen for the balance.

Diffusion Maneuver

The maneuver consists of having the patient breathe normally and then forcefully exhale to residual volume. At residual volume, the test gas is forcefully inhaled to total lung capacity, held for approximately 10 seconds, and then exhaled.

The breath-holding time should be 10 seconds, plus or minus 2 seconds. Although there are three different ways to measure the breath-holding time, the most common is for it to be measured from the beginning of inspiration of the test gas through the beginning of sample collection. The best technique for breath-holding time is probably the approach taken by Jones and Meade,1 in which the breath-hold time equals the time starting from 30% of the inspiratory time to the middle of sample collection. This technique helps compensate for slow inspiratory times that can falsely lower the Dlco, as there would be less time for alveolar diffusion.

Upon exhalation of the sample gas, before the alveolar sample is collected, a washout volume should be allowed to escape into the atmosphere. The washout volume would consist of inspired gas that had been inhaled late and that had resided, during the 10-second breath hold, in the large airways and in the larynx/pharynx (and would therefore not take part in gas exchange). For a typical patient with a vital capacity of more than 3.5 L, the washout volume should be 1 L and the following sample volume, which represents expired gas from the alveoli, should be 0.5 L. If a continuous analyzer with graphical displays is used, a post-test visual inspection of the expired carbon dioxide and tracer-gas curves can be used to adjust washout volumes. If post-test adjustments in washout volumes and points of analysis are made, caution must be used, as final diffusion results can be changed dramatically.

Calibration Techniques

Calibration of volumes in preparation for the Dlco measurement is usually done using a pressure-differential pneumotachometer. Volume calibration using a 3-L syringe should be performed at least once per day. The usual standard for calibrating a pressure-differential pneumotachometer is to inject a 3-L syringe at slow, medium, and fast speeds. For each injection speed, the resulting volume should be 3 L plus or minus 2%. Before each patient test, a two-point gas calibration should be performed for both carbon dioxide and methane or helium (whichever is used for the test).

Dlco is reported in mL per minute per mm Hg. The primary equation for calculating Dlco (where Va = alveolar volume; t = time; Ln = natural logarithm; FA = alveolar fraction; coi = carbon monoxide, initial; cof = carbon monoxide, final; Vi = volume inspired; Hef = helium, final; and Hei = helium, initial) is:

Dlco = VAx60/(Pb–47)xt+LNFACOi/FACOf
FACOf = VIxHef/HeixCO VA = VIxHei/Hef

In this calculation, the factors t and 60 represent the 10-second breath hold. The Pb is a factor in helping to drive or push the carbon dioxide across the alveolar capillary membrane. The Facoi represents the amount or fraction of carbon dioxide inhaled as the test gas, and Facof is the fraction expired in the alveolar sample.

The second equation is a single-breath helium dilution to calculate VA without the subtraction of anatomical dead space. The ratio of Facoi/Facof indicates the amount of carbon dioxide actually diffused as a result of the single-breath maneuver. Since it represents the carbon dioxide diffused during the entire 10-second breath hold, during the beginning of the breath hold, there would be a faster rate of diffusion, since there is initially more carbon monoxide in the alveoli. The Ln has the function of producing a constant rate of diffusion from the beginning to the end of the breath-hold time. To take the Ln of a function is to obtain a constant rate from a nonlinear process.

During the Dlco maneuver, the 10% helium or 0.3% methane in the test gas enables a single-breath dilution to obtain the Va. The Facoi/Facof indicates the percentage of helium that actually diffuses (or the rate of carbon dioxide transfer and uptake by the Hb).

Looking at the equation in general, the Dlco is based on the Va added to the carbon dioxide percentage of uptake by the blood. This means that the larger the Va, in general, the larger the Dlco, when the carbon dioxide uptake is constant. The larger the uptake of carbon dioxide, the larger the Dlco. People with bigger lungs have higher rates of diffusion, so to correct for Va or lung volume, one divides the Dlco by the Va to obtain the Dlco corrected for lung volume.

The Dlco/Va is a useful tool for patients with normal or hyperinflated lungs (as seen in emphysema). For patients having interstitial lung disease or other types of restrictive disease, Dlco/Va will give a falsely high reading, as the actual condition of the alveolar capillary membrane may be low, but the Dlco is adjusted upward to correct for the low lung volume.

The percentage of carbon dioxide uptake by the blood passing through the capillaries is determined by analyzing the ratio of the fraction of alveolar-inspired carbon dioxide divided by the fraction of alveolar-expired carbon dioxide:


The Facof is calculated using the following equation:

Facof = VIHef/Heixco

Again, 0.3% methane may be used as a substitution for 10% helium. This calculation basically represents the ratio of carbon dioxide inhaled over the alveolar carbon dioxide collected upon exhalation. According to the primary equation, the larger the ratio, the larger the Dlco.

When performing Dlco, the technologist must obtain at least three trials. If more than four trials are done, than there probably will be buildup of carbon monoxide back pressure in the blood, resulting in lower successive Dlco results. The criterion for repeatability is the requirement that at least two acceptable tests should be within 3 mL of carbon monoxide, or that each should be within 10% of the highest value.

It was found, in a university-based laboratory study2 of normal subjects, that intrasubject variation was 3.1% This variation is a reasonable standard to follow in most laboratories. When two trials are obtained within the 3% variation, they should be averaged to be reported. If only one trial is chosen for reporting due to difficulty in testing, then a note to that effect should be included in the report.

An additional consideration that can have an effect on Dlco results is how the breath hold is maintained. On most devices, after the patient inspires to total lung capacity with the test gas, a manual or automatic valve closes, so that the patient’s held breath rests against the valve. Another type of device might allow the patient to maintain the breath hold by resting against a closed glottis, or by maintaining the breath hold with no help at all. It has been found that maintaining the breath hold with no valve or by not resting against a closed glottis can result in a Dlco up to 4 mL of carbon monoxide higher.

Dlco values can depend on a number of physiological factors, including age, gender, height, and race. The final measurement can be affected by Hb, lung volume, COHb, and Pio2 (altitude). Predicted values should be based only on age, gender, and height. There are many predicted equations for men, women, and children, but optimally, the predicted equation should be based on normal values produced in the testing laboratory.

For accuracy, at least 200 subjects of each age group should be tested before predicted equations are calculated. I would recommend that only nonsmokers be included, and the subjects should be relatively healthy. Of course, the task of developing one’s own predicted values for Dlco is very time-consuming and is beyond the capabilities of all but a few laboratories. If it is impossible, when choosing among sets of published predicted values, try to select those based on patient populations that closely match the patients in the local area. Test a number of healthy patients, perhaps approximately 20, to see whether the population’s actual values match the predicted values.

Altitude needs to be considered when choosing predicted values for Dlco; at altitude, the Pao2 will be lower, and Dlco will change by approximately 0.35% per mm Hg change in Pao2, or by 0.31% per mm Hg decrease in Pio2. When choosing predicted equations for a laboratory at sea level, predicted values drawn from patients in Salt Lake City or Denver would not be appropriate. The equation for adjusting the predicted Dlco for altitude is:

Dlco (predicted for altitude) = Dlco predicted/(1+0.0031[Pio2–150])

Hb, COHb, and exercise all can affect the actual Dlco results. Hb is an important factor in carbon monoxide transfer because the Dlco result is dependent upon the availability of Hb to bind. The standard Hb value in adults is 14.6 g per dL. In case other values for Hb are present, the predicted values, and not the actual values, should be adjusted. A general rule in pulmonary function testing is that only predicted values should be changed, not the actual patient values. The equation for Dlco Hb corrections for adolescents and adult males is:

Dlco (predicted for specific Hb) = Dlco predictedx(1.7Hb/(10.22+Hb))

It can be safely stated that Dlco is the most exacting test that we perform in the pulmonary function laboratory, and its successful completion can be considered an art because there are so many factors that must be considered and so many steps that need to be taken in preparation.

James A. Harvey, MS, RPFT, works in the Pulmonary Physiology Laboratory, Stanford Hospital and Clinics, Palo Alto, Calif.


  1. Jones RS, Meade FA. Theoretical and experimental analysis of anomalies in the estimation of pulmonary diffusing capacity by the single breath method. QJ Exp Physiol Cogn Med Sci. 1961;46:131-43.
  2. MacIntyre R, Crapo RO, Viegi G, et al. Standardisation of the single-breath determination of carbon monoxide uptake in the lung. Eur Respir J. 2005;26:720-35.