Effective determination of pre-chronic obstructive pulmonary disease by symptoms and CT features: a multicenter cross-sectional study

Introduction

Chronic obstructive pulmonary disease (COPD) is a heterogeneous pulmonary disease characterised by chronic respiratory symptoms (cough, expectoration, and dyspnoea) and persistent, progressive airflow obstruction resulting from an abnormal airway (bronchitis and bronchiolitis) and alveolar (emphysema) (1). The overall prevalence of spirometry-defined COPD was 8.6% [95% confidence interval (CI): 7.5–9.9%], accounting for 99.9 (95% CI: 76.3–135.7) million people with COPD in China, according to the latest research (2). Epidemiological data indicate that COPD is the leading cause of morbidity, mortality, and healthcare use worldwide (3).

People with COPD experience a progressive loss of pulmonary function, and once diagnosed, modifying the disease progression is difficult (4). Rather than an accelerated decline rate of forced expiratory volume in 1 second (FEV₁) in the more severe disease, a decline in pulmonary function occurs at a faster rate in the early stages of the disease (5). If the disease is identified early, there is a significant potential for intervention (6). Consequently, there has recently been a general call for efficient identification of high-risk patients with COPD to improve the prognosis of the disease.

Global Initiative for Chronic Obstructive Lung Disease (GOLD) 2022 first proposed the term ‘pre-COPD’, drawing from the concept of ‘pre-disease’ status adopted by other medical disciplines such as pre-diabetes, pre-hypertension, pre-cancer, or pre-eclampsia. Pre-disease in these disciplines does not mean that all of the subjects will develop the disease; however, it classifies a population particularly at risk for the disease and needs closer follow-up and risk management. The pre-COPD phase refers to a period of identification of individuals of any age with respiratory symptoms, with or without detectable structural and/or functional abnormalities, who, in the current absence of airflow obstruction, may develop persistent airflow limitation over time, that is, COPD (7). The prevalence of pre-COPD without airflow obstruction in the Spanish general population has been reported to be 22.3% (8), which suggests that the general population in pre-COPD is numerous. Therefore, it is of great significance to effectively identify patients with “pre-COPD”.

However, at present, it is difficult to establish an objective and comprehensive evaluation of “pre-COPD” with a clear definition of “respiratory symptoms” and “detectable structural or functional abnormalities”. Consequently, researchers have proposed the concept of GOLD 0 to help identify patients with pre-COPD (9). GOLD 0 is defined as a stage in which risk factors (smoking) and symptoms (chronic cough and phlegm) are present. FEV₁/forced vital capacity (FEV₁/FVC) is equal to or greater than 0.7. Similarly, nonobstructive chronic bronchitis (NOCB), another specific subtype of pre-COPD, is characterised by chronic cough and sputum production for at least 3 months, persisting for at least 2 years, with FEV₁/FVC exceeding 0.7 (10). However, these subtypes identify people with pre-COPD based solely on their symptoms, covering only a small subset of those who eventually progress to COPD (9). In 2022, GOLD introduced the concept of preserved ratio impaired spirometry (PRISm) as a newly identified subtype of pre-COPD, defined as pulmonary function conforming to FEV₁/FVC ≥0.7 and FEV₁ <80%pred (9). This subtype emphasizes functional abnormalities in patients with pre-COPD. However, many knowledge gaps regarding this category persist, as many of these individuals will eventually exhibit normal spirometry (11,12).

Recent studies have identified spirometric small airway dysfunction (SAD) as an important functional abnormality in pre-COPD and defined another subtype of pre-COPD (13,14). This subtype takes into account both symptoms and functional abnormalities of pre-COPD, encompassing patients with chronic cough, sputum, and spirometric SAD, yet maintaining FEV₁/FVC ≥0.7. Notably, a 5-year follow-up study reported that 27.8% of such patients developed COPD during the follow-up period compared with 6.6% in the PRISm group (13). In addition, in China, the latest China Lung Health study reported prevalence rates of 5.5% for PRISm and 7.2% for pre-COPD (defined as patients with respiratory symptoms and spirometric SAD, but FEV₁/FVC ≥0.7) among 50,991 adults over the age of 20 years, respectively (14). This finding suggests that the phenotype characterised by spirometric SAD addresses the limitation that PRISm cannot eventually progress to COPD, thereby broadening the scope of identifying patients with pre-COPD. Therefore, it is more efficient to identify pre-COPD patients with the definition of “chronic cough and sputum, pulmonary function indicating SAD, and FEV₁/FVC ≥0.7”.

Notably, as chest computed tomography (CT) is an important tool for lung cancer screening, it has been included in the routine physical examination of Chinese residents (15). We suppose that chest CT may serve as an adjunct tool for pre-COPD identification in symptomatic patients with accessible CT data, particularly when spirometry is unavailable.

However, there is a scarcity of CT studies focusing on the identification of pre-COPD. Leveraging quantitative analysis of lung parenchyma, bronchi, and pulmonary blood vessels in chest CT, this study aimed to construct a pre-COPD diagnostic model based on the regression of functional small airway disease (fSAD) dysfunction. We hypothesised that identifying pre-COPD among patients already have undergone chest CT may add clinical value to existing imaging resources. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1354/rc).

Methods

Characteristics of participants

From August 2021 to August 2024, 405 patients with chronic cough and expectoration (for at least 3 months and lasting for at least 2 years), who have undergone screening chest CT, at the Physical Examination Centre and Respiratory Department of Wuxi People’s Hospital and Community Health Service Center (including Wuxi Ninth People’s Hospital in the west of Wuxi and Xinan Hospital in the east of Wuxi) were included in this study. According to the pulmonary function results, the patients were divided into three groups: FEV₁/FVC ≥0.7 and normal small airway function group, FEV₁/FVC ≥0.7 and abnormal small airway function group, and FEV₁/FVC <0.7 (COPD) group. Forty-nine patients were excluded based on the following exclusion criteria: (I) inability to cooperate with the dual-gas phase of the chest CT examination; (II) software analysis of poor CT image quality; (III) underlying pulmonary diseases, including lung parenchyma, lung cancer, pulmonary interstitial fibrosis, pulmonary infection, thoracic spine deformity, pleural effusion, pulmonary hypertension, severe tuberculosis, bronchiectasis, and asthma; and (IV) a history of thoracic surgery. In total, 219 participants were enrolled in this study (Figure S1). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the ethics board of the Affiliated Wuxi People’s Hospital of Nanjing Medical University (No. KY24065), the ethics board of Xinan Hospital (No. 2024000), the ethics board of Wuxi Ninth People’s Hospital (No. KS24093) and the ethics board of Hudai Hospital (No. 2024-KY-001). All patients provided written informed consent to participate and publish.

Questionnaire survey

The questionnaire included questions on age, sex, weight, height, and smoking status. Body mass index (BMI) was calculated by dividing the weight by the height squared (kg/m²). The pack-years was calculated by multiplying the number of packs smoked per year by the number of years they have smoked.

Pulmonary function tests (PFTs)

All PFTs were performed within 1 week before or after CT scanning. The procedure was performed using the German JAEGER MS-PFT instrument. We have carried out unified training for the personnel of pulmonary function examination in each center, requiring that (I) blow at the fastest speed to blow out the expiratory peak; (II) exhale to residual air level, plateau for more than 2 seconds; (III) repeat the operation more than twice, FVC variation rate <5% or absolute value <100 mL, to ensure the accuracy and consistency of the final inspection results. We included the following pulmonary function parameters: FEV₁/FVC, FEV₁%, maximal mid-expiratory flow (MMEF) [forced expiratory flow (FEF) 25–75%], and maximum expiratory flow (MEF) 25, 50, and 75. We assessed SAD using two of the three spirometry indices MMEF, FEF50%, and predicted FEF75% <65%, as reported in previous studies (16,17).

CT scanning

All patients underwent breath-hold training before scanning to make sure they held their breath at the end of a deep inhale and exhale. CT scans of the inspiratory and expiratory chest were performed in all patients using the 128-slice helical CT scanner (SOMATOM go.Fit and SOMATOM perspective 128). After completing the CT scans of the thorax, one radiologist evaluated the image quality and excluded the image data that produced respiratory artifacts.

Image analysis

The raw Dicom data of the CT images were transferred to the workstation (Dexin FACT-Digital Lung workstation, Xi’an, China) for parametric response mapping (PRM) analysis. The voxels were divided into three categories according to CT values on paired respiratory CT images, which were: (I) emphysema; voxels less than or equal to −950 HU on the inspiratory image and less than −856 HU on the expiratory image; (II) fSAD; voxels greater than −950 HU on the inspiratory image and less than or equal to −856 HU on the expiratory image; and (III) normal lung; voxels greater than −950 HU on the inspiratory image and greater than −856 HU on the expiratory image. We calculated the volume percentage of each voxel category (PRM^empha%, PRM^fSAD%, PRM^normal%) (18) (Figure S2). Airways were automatically segmented, and medium-size airway parameters, including the mean wall thickness (WT), lumen diameter (LD), and wall area ratio (WAR) of the fifth-generation bronchus, were recorded. To measure the small pulmonary vessels, we restructured the pulmonary vascular vessels and selected three levels for counting. The upper cranial slice was taken approximately 1 cm above the upper margin of the aortic arch, middle slice approximately 1 cm below the carina, and lower caudal slice approximately 1 cm below the right inferior pulmonary vein (19). These CT images were analysed using a semiautomatic image-processing program. Cross-sectional area (CSA) <5 referred to the total CSA of vessels <5 mm² of the three slices, and %CSA <5 referred to the percentage of the CSA <5 to the total lung area in the selected slices. The above parameters were calculated at the level of the whole, left, and right lung and in each lung lobe, respectively.

Statistical analysis

Based on pilot data, we anticipated moderate effect sizes (Cohen’s d =0.5–0.8) for key CT parameters between pre-COPD and control groups. For analysis of variance (ANOVA) with three groups (α=0.05, power =0.8), 64 participants per group were required to detect an effect size of f=0.35 (medium effect). Accounting for 20% attrition, we targeted ≥77 participants per group. For multivariable logistic regression with least absolute shrinkage and selection operator (LASSO) selection, we adhered to the events per variable (EPV) criterion. With an estimated 77 cases of spirometric SAD (primary outcome), we conservatively limited the model to ≤7 predictors (EPV >10). Final enrollment (N=219) exceeded minimum requirements across all comparisons: normal 68, non-COPD with SAD 77, COPD 74. This provided >90% power to detect area under the curve (AUC) ≥0.8 in ROC analysis for pre-COPD discrimination. The Shapiro-Wilk test was used to assess the normality of continuous variables. For non-normally distributed variables, median and interquartile range (IQR) were presented, while for normally distributed variables, mean ± standard deviation (SD) were presented. Levene’s test was used to assess the homogeneity of variance across groups. ANOVA was used for normally distributed continuous variables with homogeneous variance across groups to test for differences between the groups. The Kruskal-Wallis test was used for other continuous variables. Categorical variables are presented as counts (%) and were tested using the Chi-squared test or, if appropriate, Fisher’s exact test. P<0.05 was considered statistically significant. To analyse the correlation between pulmonary function parameters and chest CT factors, Pearson’s correlation coefficients were calculated, and P values were adjusted using the Bonferroni correction.

Model construction

The entire dataset was divided into two new sets by stratified random sampling, with four-fifths (N=175) assigned as the training set and one-fifth (N=44) as the test set. We adopted a two-step strategy for dimension reduction and variable selection. First, we used the LASSO to solve collinearity problems between variables and selected the most valuable variables as candidates for further regression analysis (1). Univariate regression analysis was performed on the training set to identify the independent determinants of outcomes. Multiple comparisons were adjusted using the false discovery rate method (measured by adjusted P value) to control the overall false-positive rate at the 5% level (2). Variables with adjusted P<0.05 in univariate analysis were introduced into a multivariate model. Then, variance inflation factors (VIFs) were calculated to test the collinearity between variables. The R software, version 4.2.2 (glmnet package), was used to perform the analysis.

Results

Comparison of demographic, PFT characteristics, and CT parameters among the three groups

A total of 219 participants were grouped, based on the PFT results, into normal participants (n=68), non-COPD with SAD participants (n=77), and COPD patients (n=74). Significant differences in the demographic characteristics, except height (P<0.05), were observed, as presented in Table 1. Table 2 (including CT parameters of the whole lung, left lung and right lung) and Table S1 (including CT parameters of each lung lobes) present the significant differences in the PRM values and parts of the airway (mainly based on the parameters of the right lung and right upper lobe) and vascular parameters (mainly based on the percentage of the pulmonary vessel area). We discovered that the difference in the PRM values between the three groups was more significant than that in the airway parameters, and the difference in the vascular parameters seemed to be the least significant; however, it is still clear that the small blood to total pulmonary vessel volume ratio decreased as the disease progressed.

Correlation between CT parameters and PFT characteristics

The results are presented in Table 3 (including CT parameters of the whole lung, left lung and right lung) and Table S2 (including CT parameters of each lung lobes). Among them, 22 CT parameters significantly correlated with FEV₁%, 25 with FEV₁/FVC, 28 with MMEF75/25, 27 with MEF50, and 23 with MEF75. A few medium-sized airway values showed a statistically significant but weak correlation with FEV₁% and FEV₁/FVC (P<0.05, r<0.3). Some small pulmonary vessel values showed statistically significant positive correlations with MMEF75/25 and MEF50. However, these correlations were weak (r<0.3). Compared with medium-sized airway values and small pulmonary vessel values, PRM values were more strongly correlated with pulmonary function results, and the correlation was moderate (r=0.3–0.6, P<0.05).

Construct regression models of SAD for early detection of COPD

We finally selected age, weight, smoke, packyear, ur_LD, ul_WAR, ll_WAR, all_PRM^normal, ur_PRM^fSAD, lr_PRM^fSAD as input variables for the regression model of SAD (Figure S3A, Tables 4,5), and the receiver operating characteristic (ROC) curves of the single variables against the model (Figure 1A). The AUC was used to assess the model discrimination. In the training set, the AUC was 0.9146 (95% CI: 0.8573–0.9718) (sensitivity 0.8852, specificity 0.8182) and 0.9421 (95% CI: 0.8574–1) in the validation set.

Figure 1 ROC curve of univariate and multivariate models. (A) ROC curve of the univariate and multivariate analyses in the training set in the regression model of spirometric small airway dysfunction. The AUC was used to assess the model discrimination. In the training set, the AUC was 0.9146 (95% CI: 0.8573–0.9718) (sensitivity 0.8852, specificity 0.8182), which had good significance. (B) The ROC curves of the univariate and multivariate analyses in the training set in the regression model of COPD. The AUC of the multivariate model was 0.8662 (95% CI: 0.7902–0.9422) (sensitivity 0.9091, specificity 0.7857). All, whole lung; AUC, area under the curve; l, left lung; CI, confidence interval; COPD, chronic obstructive pulmonary disease; LD, lumen diameter; ll, left lower lung; lr, right lower lung; PRM, parametric response mapping; SAD, small airway dysfunction; ROC, receiver operating characteristic; ul, left upper lung; ur, right upper lung; WAR, wall area ratio.

The same method was used to discriminate between patients with and without COPD model construction. The predictors were age, l_PRM^normal and all_PRM^empha (Figure S3B, Tables S3,S4). The AUC of the multivariate model was 0.8662 (95% CI: 0.7902–0.9422), (sensitivity 0.9091, specificity 0.7857) (Figure 1B) and 0.9583 in the validation set (95% CI: 0.8921–1). Figure 2 shows the nomogram of each model.

Figure 2 Nomogram of the regression models. (A) Nomogram of the regression model of spirometric small airway dysfunction. (B) Nomogram of the regression model of COPD. All, whole lung; COPD, chronic obstructive pulmonary disease; l, left lung; LD, lumen diameter; ll, left lower lung; lr, right lower lung; PRM, parametric response mapping; SAD, small airway dysfunction; ul, left upper lung; ur, right upper lung; WAR, wall area ratio.

Discussion

In this study, we analysed the correlations between the quantitative CT and pulmonary function parameters and constructed a regression model of SAD using the quantitative CT characteristics that allows the identification of pre-COPD.

The general population in the early stages of COPD is numerous. A UK biobank cohort analysis found that 38,639 (11.0%) of 351,874 participants had PRISm at the beginning of the study (20). Compared to normal spirometry, baseline PRISm was associated with a small but statistically significant increased risk of mortality and adverse cardiovascular and respiratory outcomes (21). Identifying high-risk patients in the early stages of COPD to improve the disease prognosis is necessary.

A clearer definition of pre-COPD based on the effects of potential indicators of increasing risk of COPD development and a diagnostic model involving them are lacking. In 2021, Han et al. reviewed studies on the early stage of COPD progression and proposed a set of clinically implementable thresholds to identify people with pre-COPD by combining symptoms, structural abnormalities, and functional abnormalities (9). As research progresses, the heterogeneity of patients with pre-COPD has become increasingly recognised, and many subtypes of pre-COPD have been explored. Some subtypes are characterised by a high burden of symptoms. Lambert and Bhatt defined “people with clinical symptoms and a history of smoking with normal lung function” as high-risk groups for COPD (22). In 2023, Divo et al. established a simple model for identifying high-risk individuals with chronic airflow restriction in middle-aged smokers, incorporating smoking history, BMI, chronic bronchitis, and FEV₁/FVC as predictors, and this model was also well validated in external cohorts (23). However, chest CT features were not included in the model owing to economic considerations, which may reduce the accuracy of the model and narrow the detection range of cases with chronic airflow restriction. Moreover, some subtypes of pre-COPD highlight abnormalities in lung function. Among them, PRISm has been widely utilised in the early stage of COPD research. However, a significant limitation of PRISm is that many individuals identified as PRISm do not develop COPD during follow-up (11). A recent study has reported that using FEV₁/FVC before the 10th percentile as a threshold can accurately identify individuals at high risk of developing COPD in the general population, with patients below this threshold having a 36-fold increased risk of developing the disease during an eight-year follow-up period (24). Spirometric SAD is also an important feature in pre-COPD research. One study divided 196 α1-antitrypsin deficiency (AATD) patients into those with normal FEV₁/FVC and MMEF (group 1), normal FEV₁/FVC and reduced MMEF (group 2), and spirometrically defined COPD (group 3). The results showed that patients in group 2 had worse health status than did those in group 1 and had a greater subsequent decline in FEV₁, indicating that a reduction in MMEF is an early feature of lung disease in COPD (25). Many studies have also confirmed that the narrowing and disappearance of small airways occurs before the destruction of emphysema, suggesting an early stage of COPD (26,27). In addition, a recent study indicated that individuals characterised by chronic bronchitis and SAD of lung function were more likely to progress to COPD [hazard ratio (HR) 5.89 and 4.80] (13). Therefore, it is reasonable to assume that SAD represents the early stage of COPD progression, and subtypes characterised by symptoms and SAD can be more efficient in identifying patients with pre-COPD. At the same time, we also acknowledge that there may be other subtypes whose phenotypes have not yet been defined, which will further improve our understanding of the pre-COPD population.

Current researches have revealed several promising biomarkers that might be incorporated to identify patients with pre-COPD or COPD. Machine learning models demonstrated that COPD patients can be classified from controls with >99% accuracy based solely on TRPC6 expression levels in lung tissue (28). A prospective study has showed the potential of sputum nanoparticles as markers of airway inflammation and disease, making it a tool to identify pre-COPD (29). Beyond single-gene biomarkers, volatile organic compounds (VOCs) in exhaled breath have been confirmed to be biomarkers for identifying PRISm, enabling early intervention before irreversible lung decline (30). Proteomic and metabolomic analyses further expand the biomarker landscape. MUC5AC, a gel-forming mucin, its ratio to MUC5B serves as a sensitive marker of mucus dysfunction and an increase of the ratio in smokers often indicates the onset of COPD (31). In conclusion, novel biomarkers also offer promising avenues for early COPD detection.

A total of 64 quantitative CT parameters (three medium-sized airway parameters, two vessel parameters, and three PRM parameters were evaluated, and the level of the whole, left, and right lung and each lung lobe) were included in this study. Airway abnormalities in the right lung and right upper lobe appeared to be more significant among the three groups than in other lung regions; Yasunaga et al. discovered that current asymptomatic smokers had a higher percentage of emphysema in the right lung and right upper lobe than did non-smokers (32), indicating that the CT features of the right lung and right upper lobe can better represent the structural changes in the high-risk population of COPD.

Smokers with normal lung function and individuals with mild emphysema show changes in the distal pulmonary vessels (33,34). These may represent the initial stages of COPD development. Previous studies on pulmonary vascular pruning in COPD have shown that %CSA <5 has a relatively weak association with airflow limitation but a strong association with emphysema (35,36). One study reported that pulmonary vascular pruning was more strongly associated with the emphysema phenotype than with the bronchitis phenotype in patients with COPD (37), which is consistent with our results. We hypothesised that pulmonary vascular pruning may be the result of alveolar hyperinflation and vascular compression during gas retention (38). This may explain the inconsistent relationships among peripheral pulmonary vascular alterations, airflow limitation, and emphysema.

The LASSO regression method was used to eliminate highly correlated variables, regardless of the effect of multicollinearity between the variables, to screen the most significantly changed parameters for modelling. When constructing the pre-COPD diagnostic model, we adopted the PRM^SAD-related parameters, while PRM^empha-related parameters were adopted in the COPD diagnostic model, suggesting that the key factor leading to airflow obstruction in early COPD is SAD rather than emphysema (27). As the disease progresses, emphysema becomes an important cause of air limitation. Small pulmonary vessel parameters were ultimately not included in the model, indicating that vascular parameter changes are not as significant as those in the PRM values (39). Further, CT features of the left lung appeared to be more discriminatory among patients with COPD. Consistent with previous findings, adding CT imaging characteristics to demographic data significantly improved the diagnostic of COPD progression (40).

There are some limitations in this study. This analysis used a post-bronchodilator FEV₁-FVC ratio of ≤0.70, cited by the Global Initiative for COPD and frequently used in large databases, which may underestimate the prevalence of COPD among younger people and overestimate its prevalence among older people. Clinical assessment parameters, such as clinical symptoms and the 6-minute walk test, were not included in our classification model, which may limit the interpretation of high-risk COPD. The correlation between the lung function test results and CT parameters was not as strong as that in previous studies, probably because of the relatively small sample size. This was a cross-sectional study of patients with a reduced MMEF, and it is unknown whether these patients developed COPD later in life. Further studies should enrol more participants and follow them up longitudinally to determine whether they develop COPD.

Conclusions

In conclusion, we confirmed that quantitative CT parameters are able to discriminate between the normal, non-COPD with SAD, and COPD groups. The model comprising age, weight, smoke, packyear, ur_LD, ul_WAR, ll_WAR, all_PRM^normal, ur_PRM^fSAD, lr_PRM^fSAD could distinguish the non-COPD with SAD group from the normal group, while age, l_PRM^normal and all_PRM^empha were predictors between the non-COPD with SAD group and the mild to moderate COPD group. This study demonstrated that the utility of quantitative CT for stratifying pre-COPD among symptomatic patients, which may provide additional diagnostic value for the group of patients with chronic cough and expectoration through existing CT scans.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1354/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1354/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1354/prf

Funding: This study was supported by the Cohort and Clinical Research Program of Wuxi Medical Center, Nanjing Medical University (No. WMCC202315), the Project of Major Program of Wuxi Medical Center, Nanjing Medical University (No. WMCM202309) and Basic Scientific Research Business Fund Project at Central Universities (No. 22520231033).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1354/coif). All authors declare that the present manuscript received the support from the Cohort and Clinical Research Program of Wuxi Medical Center, Nanjing Medical University (No. WMCC202315), the Project of Major Program of Wuxi Medical Center, Nanjing Medical University (No. WMCM202309) and Basic Scientific Research Business Fund Project at Central Universities (No. 22520231033). The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the ethics board of the Affiliated Wuxi People’s Hospital of Nanjing Medical University (No. KY24065), the ethics board of Xinan Hospital (No. 2024000), the ethics board of Wuxi Ninth People’s Hospital (No. KS24093) and the ethics board of Hudai Hospital (No. 2024-KY-001). All patients provided written informed consent to participate and publish.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Venkatesan P. GOLD COPD report: 2024 update. Lancet Respir Med 2024;12:15-6. [Crossref] [PubMed]
2. de Oca MM, Perez-Padilla R, Celli B, et al. The global burden of COPD: epidemiology and effect of prevention strategies. Lancet Respir Med 2025;13:709-24. [Crossref] [PubMed]
3. Moll M, Silverman EK. Precision Approaches to Chronic Obstructive Pulmonary Disease Management. Annu Rev Med 2024;75:247-62. [Crossref] [PubMed]
4. Ritchie AI, Donaldson GC, Hoffman EA, et al. Structural Predictors of Lung Function Decline in Young Smokers with Normal Spirometry. Am J Respir Crit Care Med 2024;209:1208-18. [Crossref] [PubMed]
5. Decramer M, Cooper CB. Treatment of COPD: the sooner the better? Thorax 2010;65:837-41. [Crossref] [PubMed]
6. Doña E, Reinoso-Arija R, Carrasco-Hernandez L, et al. Exploring Current Concepts and Challenges in the Identification and Management of Early-Stage COPD. J Clin Med 2023;12:5293. [Crossref] [PubMed]
7. Agustí A, Celli BR, Criner GJ, et al. Global Initiative for Chronic Obstructive Lung Disease 2023 Report: GOLD Executive Summary. Arch Bronconeumol 2023;59:232-48. [Crossref] [PubMed]
8. Cosío BG, Casanova C, Soler-Cataluña JJ, et al. Unravelling young COPD and pre-COPD in the general population. ERJ Open Res 2023;9:00334-2022. [Crossref] [PubMed]
9. Han MK, Agusti A, Celli BR, et al. From GOLD 0 to Pre-COPD. Am J Respir Crit Care Med 2021;203:414-23. [Crossref] [PubMed]
10. Wu F, Zheng Y, Zhao N, et al. Clinical features and 1-year outcomes of chronic bronchitis in participants with normal spirometry: results from the ECOPD study in China. BMJ Open Respir Res 2023;10:e001449. [Crossref] [PubMed]
11. Wijnant SRA, De Roos E, Kavousi M, et al. Trajectory and mortality of preserved ratio impaired spirometry: the Rotterdam Study. Eur Respir J 2020;55:1901217. [Crossref] [PubMed]
12. Wan ES. The Clinical Spectrum of PRISm. Am J Respir Crit Care Med 2022;206:524-5. [Crossref] [PubMed]
13. Fan J, Fang L, Cong S, et al. Potential pre-COPD indicators in association with COPD development and COPD prediction models in Chinese: a prospective cohort study. Lancet Reg Health West Pac 2023;44:100984. [Crossref] [PubMed]
14. Lei J, Huang K, Wu S, et al. Heterogeneities and impact profiles of early chronic obstructive pulmonary disease status: findings from the China Pulmonary Health Study. Lancet Reg Health West Pac 2024;45:101021. [Crossref] [PubMed]
15. Xia C, Basu P, Kramer BS, et al. Cancer screening in China: a steep road from evidence to implementation. Lancet Public Health 2023;8:e996-e1005. [Crossref] [PubMed]
16. Xiao D, Chen Z, Wu S, et al. Prevalence and risk factors of small airway dysfunction, and association with smoking, in China: findings from a national cross-sectional study. Lancet Respir Med 2020;8:1081-93. [Crossref] [PubMed]
17. Xu J, Li J, Wang J, et al. The effects of temperature variation on obstructive pulmonary dysfunction and small airway dysfunction in asthmatic children: A continuous eight-year study in Jinan, China. Environ Res 2025;272:121096. [Crossref] [PubMed]
18. Galbán CJ, Han MK, Boes JL, et al. Computed tomography-based biomarker provides unique signature for diagnosis of COPD phenotypes and disease progression. Nat Med 2012;18:1711-5. [Crossref] [PubMed]
19. Matsuoka S, Washko GR, Yamashiro T, et al. Pulmonary hypertension and computed tomography measurement of small pulmonary vessels in severe emphysema. Am J Respir Crit Care Med 2010;181:218-25. [Crossref] [PubMed]
20. Higbee DH, Granell R, Davey Smith G, et al. Prevalence, risk factors, and clinical implications of preserved ratio impaired spirometry: a UK Biobank cohort analysis. Lancet Respir Med 2022;10:149-57. [Crossref] [PubMed]
21. Wan ES, Balte P, Schwartz JE, et al. Association Between Preserved Ratio Impaired Spirometry and Clinical Outcomes in US Adults. JAMA 2021;326:2287-98. [Crossref] [PubMed]
22. Lambert AA, Bhatt SP. Respiratory symptoms in smokers with normal spirometry: clinical significance and management considerations. Curr Opin Pulm Med 2019;25:138-43. [Crossref] [PubMed]
23. Divo MJ, Liu C, Polverino F, et al. From pre-COPD to COPD: a Simple, Low cost and easy to IMplement (SLIM) risk calculator. Eur Respir J 2023;62:2300806. Erratum in: Eur Respir J 2023;62:2350806. [Crossref] [PubMed]
24. Tan DJ, Lodge CJ, Walters EH, et al. Can We Use Lung Function Thresholds and Respiratory Symptoms to Identify Pre-Chronic Obstructive Pulmonary Disease? A Prospective, Population-based Cohort Study. Am J Respir Crit Care Med 2024;209:1431-40. [Crossref] [PubMed]
25. Stockley JA, Ismail AM, Hughes SM, et al. Maximal mid-expiratory flow detects early lung disease in α(1)-antitrypsin deficiency. Eur Respir J 2017;49:1602055. [Crossref] [PubMed]
26. Koo HK, Vasilescu DM, Booth S, et al. Small airways disease in mild and moderate chronic obstructive pulmonary disease: a cross-sectional study. Lancet Respir Med 2018;6:591-602. [Crossref] [PubMed]
27. Verleden SE, Hendriks JMH, Snoeckx A, et al. Small Airway Disease in Pre-Chronic Obstructive Pulmonary Disease with Emphysema: A Cross-Sectional Study. Am J Respir Crit Care Med 2024;209:683-92. [Crossref] [PubMed]
28. Dhong KR, Lee JH, Yoon YR, et al. Identification of TRPC6 as a Novel Diagnostic Biomarker of PM-Induced Chronic Obstructive Pulmonary Disease Using Machine Learning Models. Genes (Basel) 2023;14:284. [Crossref] [PubMed]
29. Freund O, Rotem-Green M, Rahat M, et al. Nanoparticles in induced sputum - a window to airway inflammation among active smokers. Nanomedicine (Lond) 2023;18:303-15. [Crossref] [PubMed]
30. Tian J, Zhang Q, Peng M, et al. Exhaled volatile organic compounds as novel biomarkers for early detection of COPD, asthma, and PRISm: a cross-sectional study. Respir Res 2025;26:173. [Crossref] [PubMed]
31. Radicioni G, Ceppe A, Ford AA, et al. Airway mucin MUC5AC and MUC5B concentrations and the initiation and progression of chronic obstructive pulmonary disease: an analysis of the SPIROMICS cohort. Lancet Respir Med 2021;9:1241-54. [Crossref] [PubMed]
32. Yasunaga K, Chérot-Kornobis N, Edmé JL, et al. Emphysema in asymptomatic smokers: quantitative CT evaluation in correlation with pulmonary function tests. Diagn Interv Imaging 2013;94:609-17. [Crossref] [PubMed]
33. Saruya S, Yamashiro T, Matsuoka S, et al. Decrease in Small Pulmonary Vessels on Chest Computed Tomography in Light Smokers Without COPD: An Early Change, but Correlated with Smoking Index. Lung 2017;195:179-84. [Crossref] [PubMed]
34. Bhattarai P, Lu W, Gaikwad AV, et al. Arterial remodelling in smokers and in patients with small airway disease and COPD: implications for lung physiology and early origins of pulmonary hypertension. ERJ Open Res 2022;8:00254-2022. [Crossref] [PubMed]
35. Park SW, Lim MN, Kim WJ, et al. Quantitative assessment the longitudinal changes of pulmonary vascular counts in chronic obstructive pulmonary disease. Respir Res 2022;23:29. [Crossref] [PubMed]
36. Tang G, Wang F, Liang Z, et al. Correlations of Computed Tomography Measurement of Distal Pulmonary Vascular Pruning with Airflow Limitation and Emphysema in COPD Patients. Int J Chron Obstruct Pulmon Dis 2022;17:2241-52. [Crossref] [PubMed]
37. Matsuoka S, Washko GR, Dransfield MT, et al. Quantitative CT measurement of cross-sectional area of small pulmonary vessel in COPD: correlations with emphysema and airflow limitation. Acad Radiol 2010;17:93-9. [Crossref] [PubMed]
38. Rahaghi FN, Argemí G, Nardelli P, et al. Pulmonary vascular density: comparison of findings on computed tomography imaging with histology. Eur Respir J 2019;54:1900370. [Crossref] [PubMed]
39. Pu Y, Zhou X, Zhang D, et al. Quantitative Assessment Characteristics of Small Pulmonary Vessel Remodelling in Populations at High Risk for COPD and Smokers Using Low-Dose CT. Int J Chron Obstruct Pulmon Dis 2024;19:51-62. [Crossref] [PubMed]
40. Elbehairy AF, Marshall H, Naish JH, et al. Advances in COPD imaging using CT and MRI: linkage with lung physiology and clinical outcomes. Eur Respir J 2024;63:2301010. [Crossref] [PubMed]