A27-02 Lung Function Classification Using Race-neutral Versus Race-specific Spirometry Equations in Mexican American Children With Asthma

Rationale Race-specific spirometry equations include race or ethnicity whereas the Global Lung Function Initiative (GLI) “Global” equations remove race and ethnicity from lung function predictions. Evidence is limited on how these equations affect lung function interpretation and their association with asthma outcomes among Mexican American children. Our objective is to compare the impact of race-neutral (GLI Global) and race-specific (Hankinson) spirometry equations on lung function classification and to evaluate their association with asthma morbidity in Mexican American children. Methods We analyzed data from 407 children aged 8-17 years with current asthma and acceptable-quality spirometry from the National Health and Nutrition Examination Survey (2007-2012) who self-identified as Mexican American (n = 77), non-Hispanic White (n = 146), or non-Hispanic Black (n = 184). Z-scores (z) for FEV1 and FVC were calculated using Hankinson and GLI Global equations; a z-score indicates how many standard deviations observed values differ from predicted values. Lung function was classified as normal (z ≥-1.64) or impaired (z <-1.64). FEV1/FVC was analyzed as a raw measure. Survey-weighted regression models were used to estimate associations between lung function measures and asthma outcomes: fractional exhaled nitric oxide (FeNO), wheezing attacks and asthma-related emergency department (ED) visits in the past year. Results Among Mexican American children, mean (SD) zFEV1 was higher using GLI Global (0.54 [1.11]) versus Hankison (-0.11 [1.05]) equations, increasing the percent classified with normal FEV1 (90% to 96%) and FVC (94% to 99%). zFEV1, generated by either reference equation, was not associated with any asthma-related outcomes for Mexican American children. Lower FEV1/FVC among Mexican American children was associated with higher odds of a wheezing attack in the last year, and although not statistically significant, was positively associated with FeNO (OR = 1.13) and asthma-ED visits (OR = 1.57). Reclassification was minimal among White children (FEV1: 96% to 97%; FVC: 99% to 100%), and larger and opposite in direction among Black children (FEV1: 91% to 76%; FVC: 99% to 87%). Among White and Black children, zFEV1 was not associated with any asthma-related outcomes but lower FEV1/FVC was significantly associated with higher FeNO levels (Table). Conclusions In Mexican American children with asthma, generating zFEV1 and zFVC using GLI Global versus Hankinson equations classified lung function as normal in a higher percent of children. zFEV1 was not linked to asthma-related outcomes, but FEV1/FVC was associated with wheezing attacks. Compared to zFEV1, generated by either equation, FEV1/FVC may be more strongly associated with asthma-related outcomes in Mexican American children.