Abnormal predicted diffusion capacities in healthy Asians: an inequality with a solution

BACKGROUND: Asian lung volumes are 10-15% less than those of Caucasians. OBJECTIVES: To test the hypothesis that healthy Asians might be labeled as abnormal using three commonly used Caucasian-derived prediction equation estimates (PEE) of DLCO currently used. In addition, a Chinese-derived PEE of DLCO was tested to determine its validity in non-Chinese Asians. METHODS: Forty-one healthy Asians underwent DLCO testing. Controls consisted of the PEE and 12 healthy Caucasians. Measured DLCO was compared with the Miller, Knudson, Crapo and one Chinese PEE. Abnormal was defined as a DLCO <80% predicted. Gas dilution and plethysmography estimated alveolar volume. Proportions in parentheses in the results below are DLCO adjusted for alveolar volume. RESULTS: The average Asian DLCO was 25.75 +/- 5.55 ml/min/mm Hg, no different than the predicted DLCO of 25.29 +/- 5.53 seen with Chinese PEE. This was different (p < 0.01) than the predicted DLCO of 27.82 +/- 5.09, 33.66 +/- 6.29, and 31.64 +/- 5.33 for the Miller, Knudson, and Crapo equations, respectively. This resulted in 4/41 (0/41), 27/39 (2/39), 21/41 (3/41) and 1/41 (0/41) DLCO measurements being defined as abnormal using Miller, Knudson, Crapo and Chinese PEE, respectively. In Caucasians, the measured DLCO was similar to the Miller but significantly lower than the Knudson and Crapo PEE. Measured lung volumes were significantly smaller compared to predicted for the three Caucasian PEE in Asians, with no difference in Caucasians. There was no difference in measured lung volumes and Chinese PEE. CONCLUSIONS: Current Caucasian PEE for DLCO when used in healthy Asians result in an abnormal reading that is incorrect from 10 to 50% of the time. This PEE failure is related to a reduction in lung volume not accounted for. The Chinese PEE for DLCO works for non-Chinese Asians and should replace Caucasian PEE in the US in all Asians.