Influence of reference equations on the clinical identification of pulmonary obstruction in adults. Variation according to spirometric equations

Background: Reference equations may classify pulmonary obstruction differently, affecting clinical and epidemiological comparisons across population subgroups. Objective: To analyze the influence of the GLI Global, GAMLSS, and segmented equations on the classification of pulmonary obstruction and their concordance across subgroups. Method: Analytical study using secondary data (n = 16,596). Demographic and anthropometric variables were evaluated along with FEV₁, FVC, and the FEV₁/FVC ratio. Obstruction was defined using the lower limit of normal (5th percentile) according to GLI Global, GAMLSS, and a segmented model. Associations were assessed using logistic regression, and concordance was evaluated with kappa statistics. Results: The prevalence of obstruction varied by equation. GLI Global produced higher values and larger sex differences, doubling the estimates of GAMLSS and the segmented model. In adjusted models, GLI Global identified higher risk in men and differences by race/ethnicity, whereas the other two equations remained stable and showed no sex effect. Nearly half of the cases classified as obstructive by GLI were normal according to GAMLSS and the segmented model, which demonstrated near-perfect concordance between them. Conclusions: GLI Global increases obstruction diagnoses and may lead to overdiagnosis, whereas GAMLSS and the segmented model provide more reliable results for clinical practice.