Development and validation of predictive equations for spirometry in Portuguese children

Abstract

Background: There are no data on lung function reference values for Portuguese children, and the contribution from the Portuguese data set in the Global Lung Function Initiative (GLI) is scant. Objectives: We aimed to estimate new up-to-date reference values for Portuguese children by fitting a multivariable regression model to a general population sample. Further, we intended to assess the external validity of the obtained reference values and to compare them to the GLI reference values. Methods: A random sample of 858 children from 20 primary schools were screened by health questionnaire, physical examination, and spirometry. Spirometric parameters recorded were FVC, FEV1, and FEF25-75. Multiple regression models were used to derive reference equations. Results: Overall, 481 children, aged between 7 and 12 years, 267 boys (55.5%), were included. Boys had higher values for FVC and FEV1 than girls (P < .05). The strongest correlation was found for FVC with height (r = 0.71 for boys and 0.70 for girls), while the lowest correlation was observed in both sexes for FEF25-75 with age (r = 0.23). Height was the most significant predictor of FVC, FEV1, and FEF25-75 in our models. Weight and body mass index were not significant predictors for boys but had a significant effect on girls’ equations for all spirometry parameters. Compared to obtained reference equations with GLI, they performed better for FVC in boys, FEV1 in girls, and FEF25-75 in both boys and girls. Conclusion: We offer up-to-date reference values of spirometry for Portuguese children that can be used in clinical practice and research.

Keywords

• Children
• spirometry
• reference values
• lung function
• linear regression