[Between- and within-subject variability of pulmonary function. Four year follow-up of adult males]

Maximal forced expiration of 326 adult males 30 to 55 yr of age was measured four times during a follow-up of 4 yr. Acceptable results were obtained in 270 subjects three times or more. Forced vital capacity (FVC), forced expiratory volume in one second, maximal expiratory flows at 50% and 25% FVC divided by squared height (CFVC, CFEV1, CVmax50, and CVmax25, respectively) were analyzed. Between- and within-subject variance were estimated by a linear model, including the effects of individual and age as explanatory variables. The goodness of fit of the model was satisfactory, since the coefficients of determination of the model were more than 0.96 for CFVC and CFEV1 and around 0.91 for CVmax50 and CVmax25. Estimated between-subject variance was far larger than within-subject variance, being more than ten times for CFVC and CFEV1, about eight times for CVmax50, and five times for CVmax25. At the same time, annual decline of pulmonary function determined cross-sectionally was significantly larger than that determined longitudinally. These results indicated the superiority of longitudinal analysis to cross-sectional analysis in evaluating relatively small effects on pulmonary function especially in the case of chronic exposure of toxic substances in low doses.