Percentiles of normality for spirometry in Italian children aged 11-16 years

Aim: to provide a set of percentiles of normality, from 5% to 95%, to assess FEV1, FVC and FEF25-75% in children. Methods: separately in males and females, quantile regression was applied to estimate each percentile from 5% to 95% as a function of age, height, and BMI, on 667 normal children (54.4% males, age 11-16 years) selected from a sample of 3227 children drawn from 20 junior high schools of Palermo, Southern Italy. Subjects were excluded if they reported any respiratory symptom or history of chronic respiratory disease. No child was an active smoker. Each subject filled in a respiratory questionnaire and performed spirometry through a MicroMedical MicroLoop turbine spirometer. Results: both in males and females, height was the single most important and significant predictor of the observed FEV1, FVC and FEF25-75%, for all considered percentiles. Age significantly improved prediction accuracy in particular of FEV1, in both genders. BMI was a significant predictor of FVC, especially for the higher percentiles of both genders. For example, in females the 50th percentile (median) of FVC increased by 40 mL for a 1-cm increase in height, by 28 mL for 1-kg/meter2 in BMI, and by 54 mL for a 1-year increase in age. Conclusions: quantile regression provides a set of percentiles for each spirometric parameter, which may be used as a range of normality when assessing lung function measurements in any given child. The use of percentiles of normality should be encouraged, for they may provide additional diagnostic insight in lung function assessment. Funded By: ARPA Sicilia.