Relationship of lung function loss to level of initial function: correcting for measurement error using the reliability coefficient.

The regression of lung function change on the initial lung function level is biased when the initial level is measured with random error. Several methods have been proposed to obtain unbiased estimates of regression coefficients in such circumstances. We apply these methods to examine the relationship between lung function loss over 11 years and its initial level in 433 men aged about 20 when first seen. On theoretical and practical grounds the best method is the correction of the regression coefficient using the reliability coefficient. This is defined as the ratio of the error free variance to the variance of the variable measured with error, and is easily estimated as the correlation between repeat measurements of the underlying level. In young men the loss of some lung functions (forced vital capacity [FVC], forced expiratory volume in one second [FEV1], forced expiratory flow in the middle half of expiration, and the ratio FEV1/FVC) do not appear to be related to initial level.