Reference Equations for Spirometry in the Canadian Population.

Abstract

Spirometry plays a major role in the diagnosis and assessment of severity of lung disease. Determining which lung function values are normal and which are below the lower limit of normal depends on reference equations derived from an appropriate population. The purpose of this study was to derive spirometric reference equations for the Canadian population. The Canadian Health Measures Survey consisted of a respiratory questionnaire, urinary cotinine measurements, and spirometry performed in the sitting position with rigorous quality control standards. Of the 16,606 respondents between 6 and 79 years of age, 11,145 were eliminated for positive responses to the respiratory questionnaire, tobacco exposure, or inability to provide high-quality spirograms. Of the remaining 5,461, roughly half were less than 18 years of age. Quantile regression was used to derive predictive (median) and lower limit of normal equations for males and females for FEV1, FVC, and FEV1/FVC ratio for those with ages greater and less than 18 years. The resulting equations were compared with those from the Global Lung Initiative (GLI) and National Health and Nutrition Examination Survey (NHANES) III by using an ideal subject on the 50th percentile for height and between the ages of 6 and 79 years; the comparison showed minor and inconsistent discrepancies among the predictive equations. A plot of residuals (predicted minus measured value for each subject) suggested a marginally better fit compared with the GLI and NHANES III equations, although differences among the equations were small and unlikely to have any clinical significance. This study provides spirometric reference equations for the Canadian population that were measured under the recommended clinical conditions and with rigorous quality control.