RT Journal Article SR Electronic T1 New paediatric spirometry reference values from the LUNOKID study: Comparison to Zapletal's and Quanjer's data JF European Respiratory Journal JO Eur Respir J FD European Respiratory Society SP P4663 VO 42 IS Suppl 57 A1 Anke Hüls A1 Ursula Krämer A1 Monika Gappa A1 Christine Müller-Brandes A1 Gabriele Seitner-Sorge A1 Andrea von Berg A1 Antje Schuster A1 Christina Beckmann A1 Sabina Illi A1 Matthias Wisbauer A1 Dietrich Berdel YR 2013 UL http://erj.ersjournals.com/content/42/Suppl_57/P4663.abstract AB Background: Spirometry reference values developed by Zapletal et al. (Zapletal A. et al., Basel: Karger 1987), although still clinically used, may no longer be appropriate. The Global Lung Initiative (GLI) (Quanjer P. et al. Eur Respir J 2012; 40; 1324-1343) developed new multi-ethnic reference values. The aim of the present analysis was to assess whether the GLI data are representative for our German population and the bias introduced if old reference equations continue to be used.Material and Methods: In our LUNOKID study, we examined 4- to 18-year-old children/adolescents in 3 German communities under field conditions. For analysis, we used the regression model LMS introduced by Quanjer et al. in the context of spirometry.Results: Of 3133 recruited healthy children 1943 had a lung function visual acceptable as well as meeting ATS/ERS criteria. Lung function did not only depend on the child’s height as supposed by Zapletal et al., but also in a non-linear way on age. Compared with the GLI data, we found that the variation coefficient in our LMS model did not depend on age. Furthermore, a significantly smaller percentage (0.9 to 1.8% depending on age) of children than the expected 5% had a FEV_1 below the lower limit of normal (LLN) predicted by Quanjer et al.Conclusions: To avoid diagnostic misconceptions, the conventional reference values according to Zapletal et al. should no longer be used. The non-linear dependence of lung function on age corresponds to the results recently published by Quanjer et al. Using the LLN as defined by the GLI data potentially leads to an underdiagnosis in a proportion of German children and adolescents.