Impacts of race-neutral GLI reference equations on spirometry interpretation in a Southeast Asian population.

BACKGROUND AND

OBJECTIVE

Transitioning from ethnic-specific (GLI-2012) to race-neutral (GLI-2022) reference equations may change the interpretation of spirometric data and the classification of ventilatory defects, but the evidence from Southeast Asian populations remain uncertain. This study evaluated the agreement between GLI-2012 and GLI-2022 interpretation frameworks in a Southeast Asian cohort.

METHODS

This cross-sectional study included 265 consecutive Southeast Asian patients who underwent pre-bronchodilator spirometry at a tertiary hospital in Vietnam. Each examination was interpreted using both reference equations. Reclassification was described with transition probabilities, agreement proportions, and prevalence- and bias-adjusted kappa (PABAK), with BCa bootstrap 95% confidence intervals. Bland-Altman method was used for quantitative agreement analysis.

RESULTS

Four-pattern concordance was high (pooled union agreement, 80.27%; multicategory PABAK of 0.854), yet 29 of 265 participants (10.9%) changed ATS/ERS pattern. GLI-2022 produced higher predicted FEV1 and FVC values and lower predicted FEV1/FVC values, yielding lower FEV1 and FVC z-scores and higher FEV1/FVC z-scores. Reclassification predominantly shifted toward possible restriction or non-specific pattern (20.75% to 27.92%), whereas obstruction decreased modestly (7.55% to 4.15%). Between-framework bias was nonconstant; at the LLN, the expected differences (95% conditional prediction intervals) for FEV1, FVC, and FEV1/FVC were -0.129 (-0.436 to 0.178), -0.229 (-0.572 to 0.115), and +0.381 (0.175 to 0.586), respectively.

CONCLUSION

In this tertiary-care cohort of Southeast Asian patients, race-neutral equations produced systematic, scale-dependent shifts and reclassification of spirometric results, particularly affecting obstruction and low-FVC patterns. Longitudinal, outcome-based validation is required to determine the clinical implications of adopting race-neutral reference equations.