Applicability of the Global Lung Initiative 2022 Reference Equations on a Sample of Healthy Adolescents in Jordan

Background/Objectives: The Global Lung Initiative (GLI) 2022 race-neutral spirometry reference equations were introduced to improve interpretability across populations; however, their performance in Middle Eastern adolescents remains insufficiently validated. This study evaluated the applicability of GLI-2022 among healthy Jordanian adolescents. Methods: Healthy adolescents were recruited from secondary schools across multiple Jordanian cities (July-November 2025). Spirometry was performed according to ATS/ERS standards using a single device and standardized procedures. GLI-2022 predicted values and z-scores were derived for forced expiratory volume in one second (FEV₁), forced vital capacity (FVC), and FEV₁/FVC. Calibration was assessed using mean (SD) z-scores and the proportion below the lower limit of normal (LLN; z < -1.645). Agreement between measured and predicted values was examined using Bland-Altman methods. LLN-based pattern classifications were compared with those obtained using the local reference equation and GLI-2012. Results: A total of 921 adolescents (482 males, 439 females; mean age 15.7-16.0 years) were included. GLI-2022 produced positive mean z-scores for FEV1 (0.51-0.73) and FVC (0.51-0.69), with low proportions below LLN for both indices (<2% in each sex), indicating underestimation of predicted lung volumes. Exact binomial testing confirmed that the observed proportions below LLN for FEV1 and FVC were significantly lower than the expected 5% in both sexes (all p < 0.001). The FEV1/FVC ratio showed smaller deviations (mean z 0.07-0.19), with 4.1% of females and 5.8% of males below LLN, and these proportions did not differ significantly from 5% (female p = 0.444; male p = 0.402). Mean observed-minus-predicted biases for FEV1 were +0.185 L in females and +0.306 L in males, and for FVC were +0.224 L and +0.351 L, respectively; FEV1/FVC bias was -0.15 percentage points in females and +0.60 percentage points in males. LLN-based pattern classification showed 98.7% overall agreement with the local equation and 99.7% with GLI-2012; concordance for obstructive and possible restrictive patterns was 93.5% and 100.0%, respectively. Conclusions: In healthy Jordanian adolescents, GLI-2022 appears to underestimate predicted FEV₁ and FVC, yielding upward-shifted z-scores and fewer volume indices below LLN, while the ratio is less affected. Although LLN-based pattern classification was largely preserved, population-specific validation remains necessary before routine clinical adoption of GLI-2022 in Jordanian adolescents; extrapolation to other Middle Eastern adolescent populations should await additional regional validation.