Race-neutral GLI-Global equations produced minimal shifts in FEV1/FVC z-scores compared to GLI-2012, with obstruction prevalence decreasing only in the Other/Mixed group (7.8% to 6.7%, p=0.012).
Observational (n=1,334)
Yes
Does the use of race-neutral GLI-Global equations compared to race-specific GLI-2012 equations alter FEV1/FVC z-scores and obstruction prevalence in an African cohort?
The transition to race-neutral GLI-Global equations introduces minimal clinical impact on FEV1/FVC z-scores and obstruction classification compared to race-specific equations in an African cohort.
p-value: p=<0.01
Abstract Rationale In 2022, the Global Lung Function Initiative (GLI) introduced race-neutral reference equations (GLI-Global) to replace traditional race-specific equations (GLI-2012). This study evaluates the impact of this change on the FEV1/FVC z-scores and the prevalence of obstruction across different racial groups. The goal is to determine whether the choice of reference equations leads to differences in obstruction prevalence and to identify potential diagnostic biases or disparities. Methods Spirometry data originated from children aged 6-14 and adults aged ≥18 years, all participants in the CHEST-Africa study investigating the prevalence of asthma and COPD in Africa. Quality of spirometry was assessed by ArtiQ.QC software and sessions graded D or lower for either FEV1 or FVC (based on ATS/ERS 2019 standard) were excluded from analysis. We analyzed spirometry data from 1334 subjects (548 Black, 765 Other/Mixed, 21 Caucasian; 693 male, 641 female; age range 6-79 years, mean 33.5). FEV1/FVC z-scores were computed using GLI-2012 and GLI-Global equations. Obstruction was defined as z −1.645. Paired comparisons of z-scores were performed using the Wilcoxon signed-rank test, distributional differences were evaluated via the Kolmogorov-Smirnov (KS) test, and changes in obstruction prevalence were tested using McNemar’s test. Results Mean FEV1/FVC z-scores were similar between GLI-2012 and GLI-Global across all races (difference0.06). Paired comparisons showed statistically significant but small shifts in z-scores for all racial groups (p 0.01). The overall distribution of z-scores did not differ significantly (KS p 0.6). Obstruction prevalence remained stable for Black (7.5%, N = 41) and Caucasian (0%) groups, while it slightly decreased in the Other/Mixed group (from 7.8%, N = 60 to 6.7%, N = 51; McNemar p = 0.012). Figure illustrates the distribution of FEV1/FVC z-scores obtained using the race-specific and race-neutral GLI equations across the three racial groups. The violin plots demonstrate the overall similarity of the two sets of equations, with only minor observable shifts in z-score distributions. Conclusions Although statistically detectable, the transition from race-specific to race-neutral GLI equations introduces minimal clinical impact on FEV1/FVC z-scores. The only meaningful difference was observed in individuals of Other/Mixed race, suggesting that the new race-neutral equations may harmonize interpretation in heterogeneous populations without substantially altering diagnostic outcomes. These findings should be interpreted with the understanding that the GLI-2012 “Black” reference equations were derived from African American cohorts, which may not fully represent African spirometric norms; nonetheless, our results support the applicability and potential equity benefits of race-neutral equations in African settings. This abstract is funded by: Clario
Rakic et al. (Fri,) conducted a observational in Asthma and COPD (n=1,334). GLI-Global race-neutral reference equations vs. GLI-2012 race-specific reference equations was evaluated on FEV1/FVC z-scores and obstruction prevalence (p=<0.01). Race-neutral GLI-Global equations produced minimal shifts in FEV1/FVC z-scores compared to GLI-2012, with obstruction prevalence decreasing only in the Other/Mixed group (7.8% to 6.7%, p=0.012).