Study of lung function data by principal components analysis.

As a rational approach to the many lung function tests available, we have subjected the results of a battery of six lung function measurements made in 458 coalminers to the statistical technique of principal components analysis. By this means the six test results were reduced to three principal components without important loss of information. The first component appeared to represent lung size and the second the degree of airflow obstruction, and the third detected impairment of gas transfer factor in excess of that explained by the first two components. The values of the first principal component, used to select men with abnormal lung function, identified more younger men with functional abnormalities than a method based on comparison of observed and predicted values of forced expiration volume in one second. The values of the second and third principal components were used to classify types of functional abnormality. It is concluded that this statistical technique provides a sensitive method of identifying men with unusual lung function, particularly younger men, in a population and can be used to define and quantify different aspects of lung function.