Restrictive spirometry versus restrictive lung function using the GLI reference values.

Background: Restrictive lung function may indicate various underlying diseases. The aim of this study was to evaluate the accuracy of different restrictive spirometry patterns (RSPs) to identify restrictive lung function (total lung capacity [TLC] < lower limit of normal [LLN]) according to reference values by the Global Lung Function Initiative (GLI) in a wide age-ranged, general population sample.

Methods: A general population sample (n = 607, age 23-72 years, smokers 18.8%) with proper dynamic spirometry and TLC measurements, was included. Accuracy of two main categories of RSP to identify TLC < LLN were evaluated: traditional RSPs (definition 1: FVC < 80% of predicted and FEV /FVC ≥ 0.7 and definition 2: FVC < LLN and FEV /FVC ≥ LLN) and RSPs defined by Youden's method (definition 3: FVC < 85.5% of predicted and FEV /FVC ≥ LLN and definition 4: FVC Z-score < -1.0 and FEV /FVC ≥ LLN).

Results: The prevalence of restrictive lung function (TLC < LLN) was 5.3%. The most accurate cut-offs for FVC to identify TLC < LLN were 85.5% for FVC% of predicted, and -1.0 for FVC Z-score. The traditional RSP definitions 1 and 2 had higher specificity (95.0% and 96.9%) but substantially lower sensitivity compared to RSP definitions 3 and 4.

Conclusion: Based on the GLI reference values, the RSP definition FVC < LLN and FEV /FVC ≥ LLN yielded the highest specificity and may appropriately be used to rule out restrictive lung function. The RSP definition with the most favourable trade-off between sensitivity and specificity, FVC < 85.5% of predicted and FEV /FVC ≥ LLN, may serve as an alternative with higher sensitivity for screening.