Abstract
BACKGROUND: Dynamic spirometry is an important investigation to differentiate between impaired and normal lung function. This study aimed to evaluate the results of lung function testing in a cohort of subjects from Northern Sweden without any known heart or pulmonary disease. Our focus was to compare with two reference materials that have showed differences in the age-dependency of lung function in Swedish subjects. METHODS: The study population consisted of 285 healthy adults (148 males, 52%) between 20-90 years of age. The subjects had been randomly selected from the population register for inclusion in a study investigating cardiac function in heart-healthy subjects, but were also assessed with dynamic spirometry. At least seven percent reported smoking. Sixteen subjects presented with pulmonary functional impairments and were excluded from the current study. The sex-specific age-dependency in lung volumes was estimated using the LMS model, where non-linear equations were derived for the mean value (M), the location (L) or skewness, and the scatter (S) or coefficient of variation. This model of the observed lung function data was compared with reference values given by the original LMS model published by the Global Lung Initiative (GLI), and with the model from the recent Obstructive Lung Disease In Norrbotten (OLIN) study, where higher reference values were presented for Swedish subjects than those given by the GLI model. RESULTS: No differences were found in the age-dependency of pulmonary function between the LMS model developed in the study and the OLIN model. Although the study group included smokers, the original GLI reference values suggested significantly lower normal values of FEV(1) (forced expiratory volume) and FVC (forced vital capacity), and consequently fewer subjects below the lower limit of normality, than both the rederived LMS and OLIN models. CONCLUSIONS: Our results are in line with previous reports and support that the original GLI reference values underestimate pulmonary function in the adult Swedish population. This underestimation could be reduced by updating the coefficients in the underlying LMS model based on a larger cohort of Swedish citizens than was available in this study.