Lung function and height in 242 8-yr-old and 299 12-yr-old children without known or suspected predisposition to lung disease were measured annually over 6 and 8 yr, respectively. Growth of forced expiratory volume in 1 s (FEV1), vital capacity, and expiratory flow after expiring 50% of vital capacity were statistically modeled by age and height by use of a multivariate normal model for longitudinal data. This method has the flexibility to fit an appropriate (not necessarily linear) mathematical description of average lung function while concurrently modeling the covariance between measures on the same individual. Differences in lung function growth between girls and boys, pre- and post-puberty, showed that girls had a steadier though less pronounced increase in lung function with height. In boys, before puberty there was deficit in lung volume relative to height (not evident in girls), which was compensated for by rapid growth after puberty. The standard error of FEV1 predictions based on current height and age were more than halved when measurements of FEV1, age, and height taken 1 yr before were incorporated. We found evidence for dysanaptic growth in childhood. Fitted models have application to early detection of departures from healthy lung function.
Meta-Analysis of Test Accuracy Studies with Multiple and Missing Thresholds: A Multivariate-Normal Model
