Implications of reference equations for interpretation of spirometry in three African countries: a cross-sectional...

The Global Lung Function Initiative (GLI) and American Thoracic Society recently endorsed a race-composite spirometry reference equation ("GLI Global"). Africa (outside North Africa) is not represented in the underlying dataset; GLI Global has not been evaluated in the region. We evaluated the fit and diagnostic implications of GLI and African (identified by scoping review) reference equations in three East/Southern African countries.

Among healthy participants from a tuberculosis household contact cohort study in Mozambique, Tanzania and Zimbabwe (age ≥10 years) with post-bronchodilator spirometry we calculated forced expiratory volume in 1 s (FEV₁), forced vital capacity (FVC) and FEV₁/FVC z-scores using different equations, the proportion of people with obstructive airways disease or preserved-ratio-impaired spirometry by different equations. We compared these measures across reference equations.

In total, 806 healthy people had good-quality post-bronchodilator spirometry. Across GLI equations, "African American" fitted best (mean±sd FEV₁ z-score -0.12±1.20, mean FVC z-score -0.35±1.19). Compared with "African American", GLI Global resulted in twice as many people being identified as having preserved-ratio impaired spirometry (22% versus 11%) with a similar proportion having obstruction (4.2% versus 3.8%). Reference equations developed in Africa conferred similar fit compared with the GLI African American equation.

Reference equations have clinical and public health implications that demand careful consideration, particularly in resource-constrained environments. Use of GLI Global may result more people being identified as having lung function impairment. Further work that includes clinical outcomes is needed to ensure that GLI Global is globally representative. The key limitation of this work is the potential for people with undiagnosed respiratory disease to have been included in the analysis.