Rationale: Single and serial spirometric data are commonly compared with predicted values to assess pulmonary function and normal lung growth.
Objectives: Do reference equations adequately describe pulmonary function in a population and in growing individuals?
Methods: We applied five sets of reference equations with appropriate age ranges to cross-sectional data of FEV(1), FVC, and FEV(1)/FVC from the United States, Estonia, and The Netherlands (1,487 boys and 1,340 girls, 6 to 18 years of age), and to serial measurements in Dutch (430 girls and 769 boys, 6 to 19 years of age) and in German and Austrian children (1,305 girls and 1,303 boys, 6 to 13 years of age).
Measurements and main results: Compared with reference equations from Polgar and Zapletal, cross-sectional FEV(1) and FVC declined between the ages of 6 and 12 and then increased, leading to a spurious change of up to 25% predicted; this pattern was most pronounced in boys. In cross-sectional data this trend was much weaker when using reference equations from Hankinson, Quanjer, and Stanojevic, and these equations provided a good fit from the age of 12 upward. In longitudinal data (i.e., within individuals), the trend was more pronounced for FEV(1) in boys than in girls. No set of equations provided a satisfactory fit in the lower limits of normal, but Hankinson and Stanojevic equations performed best.
Conclusions: Spirometric reference equations that use only height for predicting pulmonary function are unsuitable for describing the progression of pulmonary function. Those that incorporate height and age demonstrate some discrepancy with longitudinal data. Failure to take these spurious trends into account leads to significant errors in estimating the natural course of respiratory disease, in allocating patients to treatment groups, or in assessing long-term effects of drug intervention in school children and adolescents.