To the best of our knowledge, this was the largest study conducted in China, including 8,929 healthy children aged 4 to 18 from 33 research centers across 24 regions. The purpose of the study was to develop reference equations for each gender, taking into account characteristics such as age, height, and weight that could influence the variability of spirometric data.
Initially, we investigated the correlation between gender, age, height, weight, and lung function parameters. Our study found that except for FEV1/FVC and MEF75, lung function parameters of boys were higher than those of girl's, which was consistent with previous studies. Given the significant role of gender in lung function parameters, we analyzed the correlation between age, height, weight, and lung function parameters separately for different genders. As expected, regardless of gender, height was the most crucial predictors. However, in boys, the correlations with age were more pronounced than those with weight, whereas the opposite trend was observed in girls. FVC, FEV1, FEF25-75%, PEF exhibited notable positive associations with age, height, and weight among both genders. The findings of this study are consistent with those of Zheng's research [12]. A possible interpretation is that Chinese boys were higher and heavier than girls within the same age group [18].
To facilitate quick or automated interpretation of results, spirometry equipment frequently incorporates the European Zapletal predictions, which are widely accepted as the recommended standard for predicting normal lung function indices in China [7]. The Zapletal prediction equations were based on 614 healthy German children and adolescents aged 5 to 17 in 1977 [7]. In our investigation, we found that FEV1 was the best-fitting spirometric parameter from Zapletal. Nonetheless, this measure still overestimated the remaining parameters that characterize small airway performance and underestimated the FEV1/FVC ratio, with around 5% of our healthy patients being labeled as having abnormal lung function.
In 2012, the GLI released global lung function equations and multiethnic reference values for spirometry for individuals aged 3 to 95 [4]. Studies have shown that the spirometric data observed in Chinese children did not closely fit the GLI-2012 reference equations [10–12]. Furthermore, inconsistent findings have been achieved from past research; for example, one study found that the spirometric values were overestimated by the GLI prediction equations [12]. Our research aligns with recent findings from Honkong [10], which showed that the FEV1 and FVC estimates from the GLI-2012 Northeast and Southeast Asian regions tended to be underestimated, while the FEV1/FVC ratio was overestimated. These results prompt inquiries on the GLI-2012's suitability for use in various Asian locales.
The GLI global equations were created using the same GLI data set, but with adjustments to the statistical contributions of the four GLI ethnic groups to achieve equal weighting and a more valid representation of the composite GLI data [5]. The applicability of the GLI-2022 race‐neutral prediction equations has rarely been investigated in the Asian pediatric population in with such relatively large sample data [10]. And our findings revealed that the GLI 2022 reference values are not perfect for the Chinese children, with underestimations for FEV1, FVC, and FEV1/FVC ratio. Furthermore, concerns have been raised about the possibility that the racialized values could conceal modifiable risk factors for poorer lung function that are present in ethnic non-White groups [19, 20]. Further research is needed to determine whether the application of race-neutral equations would enhance lung function interpretation and, eventually, lead to better outcomes. Moreover, other small airway indexes that could aid in the treatment of asthma and the detection of COPD were not included in the GLI 2012 and GLI 2022 [21, 22].
This study was compared with earlier research on Chinese children. Overall, Y. N. Ma et al. [8], M. Jiang et al. [12], M. C. Tsai et al. [13], and Ip, M. S et al. [14] underestimated the FVC and FEV1 of healthy children, while W. T. Wang et al. [9], and Jian, W et al. [11] overestimated the FVC, FEV1, and FEV1/FVC ratio. Regarding the PEF, W. T. Wang et al. [9], Jian, W et al. [11], M. Jiang et al. [12] tended to overestimate it, and Y. N. Ma et al. [8] and M. C. Tsai et al. [13] showed underestimation of it. Among those studies, the z-score for FEV1 and FVC in equations produced by Jian, W et al. [11], showed the smallest deviation when analyzed in our sample population, and they also indicated that, besides age, sex, and height, weight was found to be an additional factor contributing to the variability of spirometric parameters in the population aging from 4 to 80. Similar to our study, the study conducted by Jian, W et al. [11], included a population from several regions in China with a relatively large population, while the other studies were only conducted in a single region of China. Large-scale data specific to Chinese children were not accessible when these equations were constructed. Moreover, parameters for MEF were rarely reported in the previous studies. Notably, there have been significant changes in the economy, lifestyle, nutrition, and living conditions over the past few decades, all of which may have substantially affected lung function [23, 24]. Therefore, the critical concern is the fact that there are no local reference equations in most regions of China, except for American and European reference values adjusted with ethnic conversion factors.
To move beyond the status quo, we created new sex-specific reference equations for predicting lung function in Chinese children aged 4 to 18. In line with global practice, our most basic model only incorporates height and age [4, 5, 7]. Independent of age and height, weight was a predictor of lung function, and adding this variable to the prediction models increased the variation explained. Therefore, in order to provide the most accurate estimation of lung function, we also give models that incorporate weight. Model 2 was discovered to be the most accurate and was therefore advised for therapeutic applications, even though both models demonstrated strong predictive accuracy. For Chinese children, these optimized reference equations present the possibility of better healthcare and more a precise diagnosis.
Among the advantages of our study are its large sample size and the inclusion of small airway indexes in addition to parameters for main airway function. These contrast with the majority of earlier investigations, which were based on relatively limited sample size and largely investigated the FEV1 and FVC. Additionally, we employed the most recent ATS/ERS recommendations, which define LLN based on statistical computation rather than a percentage, to interpret our spirometry results. Furthermore, the robustness of the predictive equations was demonstrated by the high internal validity of our obtained reference values.
There were some limitations to this study. First, this cross-sectional study, like other PFT reference studies, had limited interpretation of long-term changes in lung function. Furthermore, the healthy children in this study were characterized using a questionnaire survey and did not have blood IgE screening or a chest X-ray; thus, children with allergy disorders and lung disease were not completely excluded. These issues should therefore be considered in subsequent research.