Reference values for lung function decline in working coalminers

The main aim of the present study was to provide a set of reference equations for five- and ten-year changes in FEV, and FVC, using a large database of up to five serial lung function measurements made at five-year intervals on British coalminers between 1958 and 1978, as part of the British National Coal Board’s Pneumoconiosis Field Research (PFR).The starting point for the study was a group of 14319 miners with reliable lung function measurements, who had been examined on at least two occasions during the study period. Detailed data validation produced a final study group of 13684 men, including smokers and men with chest illnesses. Available data for these men include, as well as lung function, cumulative dust exposures, radiological category, smoking habit, and respiratory symptoms. Present analyses have focussed on a subset of 771 asymptomatic non-smokers, equivalent to 1755 man-surveys where FEV, was measured, and 1672 man-surveys where FVC was measured.Statistical analysis involved both descriptive methods and model fitting procedures. For the latter, the BMDP5V program was used, which fits a multiple regression model allowing for correlation between repeated measurements on individuals. Output includes estimates of regression coefficients and residual variance-covariance matrices.Regression analyses yielded predictive equations for FEV, and FVC in terms of age, height, weight, cumulative dust exposure and presence of simple pneumoconiosis category 2 or higher. Indicator terms for colliery of employment and survey attended were also included, as were interactions between these indicators and age. Averaging over collieries and surveys, and specializing to the case of zero dust exposure and category 0 or 1 simple pneumoconiosis, gave the desired reference equations. Illustrations of how to use the equations and associated estimates of residual variability and within-man covariance to assess (i) single measurements of lung function (ii) pairs of measurements on individual men made five years apart, are given in the text.Measurements of FEV, and FVC analysed in this study were averages (means) of results of three forced expirations, and not the maximum of three, which is the usual index nowadays. A limited analysis of a subset of the study group suggested that differences between mean and maximum were reasonably stable over age, height and smoking category. For the assessment of measurements based on the maximum of three blows, it is suggested that expected values be adjusted upward by the average differences for the subset.