Health inequities in the comorbid association between pneumoconiosis and chronic obstructive pulmonary disease: a nationwide study of age-related heterogeneity Authors: Keliang Liu, Yanhong Ren, Lifang Zhou, Xiangpei Lyu, Huanqiang Wang Published: 2026 ABSTRACT AbstractBackgroundPneumoconiosis patients have a higher prevalence of chronic obstructive pulmonary disease (COPD). While age is a known associated factor, its impact across different populations is unclear.MethodsWe included 9,964 pneumoconiosis patients from 27 Chinese provinces. Factors were identified using multivariable logistic regression, and the non-linear age–COPD association was examined with restricted cubic spline (RCS) modelling. Stratified analysis and interaction tests assessed sociodemographic heterogeneity. Sensitivity analyses comprised RCS models with varying knots and propensity score matching.ResultsThe prevalence of COPD among patients with pneumoconiosis was 24.1%. Each additional year of age was associated with a significant increase in the odds of COPD (adjusted odds ratio [aOR] = 1.03, 95% confidence interval [CI]: 1.03–1.04). RCS analysis revealed a significant nonlinear relationship, with the odds increasing more steeply at older ages (P for overall < 0.001; P for nonlinear = 0.002). Subgroup analyses showed substantial heterogeneity in these associations. Among urban residents and individuals covered by work-related injury insurance (WRII), the odds of COPD increased sharply and then gradually plateaued in later life. In contrast, among rural residents and those without WRII, the odds increased steadily with advancing age. Sensitivity analyses confirmed the robustness of these findings.ConclusionThe nonlinear relationship between age and COPD in pneumoconiosis in patients with pneumoconiosis is significantly modified by sociodemographic factors.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12931-026-03765-5.Keywords: Occupational Health, Pneumoconiosis, COPD Introduction Pneumoconiosis remains a leading occupational respiratory disease globally, contributing substantially to annual morbidity and mortality, particularly in developing countries [1, 2]. In China, it has long been the most prevalent occupational illness, pathologically defined as diffuse pulmonary fibrosis caused by occupational dust exposure [3]. Patients with pneumoconiosis often experience progressive lung function decline and have a high prevalence of serious complications, notably chronic obstructive pulmonary disease (COPD) [4]. The coexistence of these conditions is associated with accelerated clinical deterioration and increased overall disease burden. COPD, characterized by persistent airflow limitation and chronic respiratory symptoms resulting from chronic airway and parenchymal inflammation, represents a major global public health challenge [5]. However, important gaps persist in understanding the association between pneumoconiosis and COPD, especially regarding how this association is heterogeneously distributed across different subpopulations [6, 7]. Age is a well-established factor associated with both diseases [8–10]. Yet, existing studies are largely limited by the assumption of a homogeneous, linear age effect across the entire population [11, 12]. This presumption overlooks marked inter-individual differences in socioeconomic (e.g., residence and work-related injury insurance) and clinical profiles, thereby masking the underlying social determinants of the association and impeding the development of targeted public health interventions. To address these limitations, we conducted a national cross-sectional study that extends beyond conventional linear analyses. We aimed to precisely characterize the nonlinear dose–response association between age and COPD among patients with pneumoconiosis, and to systematically quantify the heterogeneity of this association across key sociodemographic and clinical subgroups. We hypothesized that the age-related association is not uniform but is significantly modified by multiple contextual factors. Our findings aim to provide actionable evidence for identifying priority subgroups and informing stratified prevention strategies. Methods Study design and participants Data for this study were derived from a nationwide cross-sectional survey conducted from December 2017 to June 2021. The survey covered 27 provinces (excluding Tianjin, Shanghai, Hainan, Tibet, Hong Kong, Macau, and Taiwan), yielding a total of 11,181 completed questionnaires [13]. Survey items included demographic and sociological characteristics, comorbidities, and clinical features. Data were collected via face-to-face interviews conducted by certified occupational disease physicians and nurses who had undergone standardized training. In some instances, patients completed questionnaires in different regions or hospitals, or at different time points within the same hospital, resulting in the exclusion of 378 duplicate questionnaires. Duplicate questionnaires were identified through logic-based verification using multiple identifiers, including ID number, name, and contact information. Additionally, 180 respondents were under medical observation: these individuals had evidence of pulmonary lesions, but the density of small nodules on chest radiographs did not meet the current criteria for stage I pneumoconiosis, requiring further medical follow-up. Furthermore, 659 additional questionnaires were excluded owing to missing key data that could not be supplemented during follow-up. After excluding these 1,217 questionnaires (378 duplicate responses, 180 participants under medical observation, and 659 with missing key data), a total of 9,964 valid questionnaires were included in the final analysis. Based on an expected COPD prevalence of 25%, with α = 0.05, 90% power (β = 0.10), and accounting for 10% loss to follow-up, the minimum required sample size was 8,900. This study enrolled 9,964 participants, exceeding this target and ensuring adequate statistical power. This study was approved by the ethics Committee. Written informed consent was obtained from all participants prior to data collection. Assessment of outcomes The primary outcome of this study was COPD. Eligible participants with pneumoconiosis were required to either hold an official occupational disease diagnosis certificate or have received a clinical diagnosis from a physician qualified in pneumoconiosis assessment. In line with previous studies [4, 13], comorbidity data were collected via questionnaire, specifically by querying participants about a history of relevant complications. COPD was identified using both post-bronchodilator spirometry and medical record documentation. According to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria, participants with a post-bronchodilator forced expiratory volume in one second/forced vital capacity (FEV1/FVC) ratio < 0.70 were classified as having COPD, regardless of any prior diagnosis recorded in their medical records [14]. All 9,964 participants underwent post-bronchodilator spirometry. Clinically approved pulmonary function testing devices were used at all centers, with routine calibration or calibration verification conducted according to manufacturers’ instructions and local quality-control requirements. In China, standardized spirometry training and quality control have been supported by the Screening and Monitoring to Improve Lung disease Early Detection (SMiLE) program. Test acceptability and repeatability followed standardized ATS/ERS spirometry criteria, and post-bronchodilator testing and COPD classification were conducted in accordance with Global Initiative for Chronic Obstructive Lung Disease (GOLD) recommendations. Data collection Demographic and sociological characteristics included age, sex, body mass index (BMI), residence(urban/rural), birth place, smoking index, alcohol intake, education level, marital status, employment status, annual personal and family income per capita, accumulated dust exposure time, work-related injury insurance, region of China, season of survey, and source of cases. BMI was calculated as weight (kg) divided by the square of height (m2). The smoking index was determined by multiplying the number of cigarettes smoked per day by the number of years of smoking [15]. Smoking index was categorized into four groups: <100, 100–199, 200–399, and ≥ 400. Comorbidities included COPD, pulmonary tuberculosis, pulmonary bullae or pneumothorax, cardiovascular, diabetes, and hypertension. Clinical characteristics encompassed type of pneumoconiosis, stage of pneumoconiosis, self-rated health, history of using immune-enhancing drugs, and history of hormone therapy for pneumoconiosis. Pneumoconiosis types were classified into three groups: silicosis, coal workers’ pneumoconiosis (CWP), and other types. Silicosis results from the inhalation of crystalline silica dust and is characterized by nodular pulmonary fibrosis, predominantly affecting the upper lobes. In contrast, CWP results from exposure to coal dust and typically presents as coal macules or progressive massive fibrosis, with more prominent involvement of the airways. Pneumoconiosis staging was divided into four categories: clinical cases, stage I, stage II, and stage III. Pneumoconiosis stage (I, II, and III) was determined according to the Chinese National Diagnostic Criteria for Pneumoconiosis (GBZ 70‑2015). Staging was based on posteroanterior high‑kilovoltage chest radiographs (CXR), which were independently interpreted by two qualified occupational disease physicians; disagreements were adjudicated by a third senior diagnostician. These national criteria were developed with reference to the International Labour Organization (ILO) International Classification of Radiographs of Pneumoconiosis [16]. Clinical cases referred to patients clinically diagnosed with pneumoconiosis who did not possess an official occupational disease certificate, primarily comprising migrant workers. Self-rated health status was evaluated using the European Quality of Life Five-Dimension Scale (EQ-5D) [17, 18], which employs a 20 cm vertical visual analog scale (VAS) ranging from 0 (worst health) to 100 (best health). Statistical analysis Categorical data were presented as counts and percentages (or proportions) and analyzed using the chi-square test or Fisher’s exact test, as appropriate. To assess the association between age and COPD, univariate and multivariate logistic regression models were constructed, with odds ratios (ORs) and 95% confidence intervals (CIs) reported. Multicollinearity was evaluated by calculating the variance inflation factor (VIF) and tolerance via a multivariate linear regression model; variables with tolerance > 0.1 and VIF < 10 were retained [19]. Additionally, fully adjusted restricted cubic spline (RCS) logistic regression models were used to assess potential nonlinear associations between age and the prevalence of COPD. Four knots were specified at the 5th, 35th, 65th, and 95th percentiles of the age distribution (42, 52, 61, and 80 years, respectively). To ensure comparability across subgroups, we used a uniform reference age of 55 years (the median age of the total study population) for all RCS models. Subgroup and interaction analyses were conducted to explore potential effect modification. To verify the robustness of the observed nonlinear trends, we conducted four sensitivity analyses. First, the RCS analyses were repeated using alternative reference ages of 45 and 50 years to assess the stability of the age–COPD association with respect to the choice of reference age. Second, models with three and five knots were compared to evaluate the influence of knot number on the overall curve shape. Third, we performed 1:1 propensity score matching (PSM) using nearest‑neighbor matching without replacement and a caliper width of 0.2. Adequate balance between groups was defined as a standardized mean difference (SMD) < 0.10. Based on the matched dataset, stratified RCS analyses with four knots were conducted to further assess the stability of the observed age‑related association patterns. The covariates included in the propensity model were selected based on the subgroup forest plot and interaction analyses. Fourth, to address potential heterogeneity introduced by the clinical cases group, we repeated the core RCS analyses after excluding these patients. To evaluate the potential for overadjustment bias, we also reran the overall RCS analysis after removing the variable representing combined pulmonary bullae or pneumothorax from the covariate set. All statistical analyses were performed using R software (version 4.3.3), and a two-sided p-value < 0.05 was considered significant. Results Baseline characteristics of participants A total of 9,964 participants were included in this study, with an overall COPD prevalence of 24.1%. The mean age was 57.9 ± 11.7 years for the total population and 56.3 ± 10.9 years for those without COPD. Among COPD cases, mean age varied by stage: 67.9 ± 12.0 years for stage I, 63.3 ± 12.6 years for stage II, 56.1 ± 10.9 years for stage III, and 57.7 ± 10.4 years for clinical cases. Significant differences were observed between groups with and without COPD in terms of age, sex, residence (urban/rural), birth place, education level, marital status, body mass index (BMI), smoking index, alcohol intake, region of China, work-related injury insurance (WRII), self-rated health (SRH), comorbid pulmonary bullae or pneumothorax, and type of pneumoconiosis (all P < 0.001; Supplementary Table 1). Prevalence of COPD and associated factors in patients with pneumoconiosis The prevalence of COPD was 19.8% among rural pneumoconiosis patients and 29.7% among urban patients. Univariate analysis indicated that several factors were associated with higher odds of COPD in these patients, including demographic and socioeconomic characteristics (age, residence, BMI, smoking index, alcohol intake, employment status, annual personal and family income per capita, and WRII), comorbidities (pulmonary bullae or pneumothorax, pulmonary tuberculosis, cardiovascular disease, diabetes, and hypertension), clinical characteristics (type and stage of pneumoconiosis, self-rated health), as well as history of using immune-enhancing drugs, and history of hormone therapy for pneumoconiosis. Specifically, smoking index > 399 (OR = 1.54), BMI < 18.5 kg/m² (OR = 2.48), stage III pneumoconiosis (OR = 1.52), and the presence of pulmonary bullae or pneumothorax (OR = 3.75) were all significantly associated with higher odds of COPD (all P < 0.05; Table 1 and Supplementary Table 2). Table 1Prevalence and Odds ratios of COPD associated with social determinants and clinical factors among pneumoconiosis patients in 27 provinces of China, stratified by urban and rural residencePrevalence (%) in patients (95%CI)OR95%CI P ItemsLiving in rural areas Living in urban areas TotalSmoking index < 10015.1 (13.8–16.6)28.3 (26.5–30.2)21.3 (20.2–22.5)1.00Reference 100–19921.4 (18.2–25.0)21.0 (17.3–25.3)21.3 (18.8–24.0)1.000.84–1.180.967 200–39923.6 (21.0-26.4)29.2 (25.7–32.9)25.8 (23.7–28.0)1.291.13–1.47< 0.001 > 39924.7 (22.7–27.0)36.4 (33.5–39.4)29.4 (27.7–31.2)1.541.38–1.72< 0.001 P for trend< 0.001< 0.001< 0.001BMI < 18.5 kg/m235.7 (31.8–39.8)39.8 (33.4–46.6)36.8 (33.4–40.4)2.482.00-3.07< 0.001 18.5–23.9 kg/m219.8 (18.5–21.2)29.1 (27.1–31.1)23.3 (22.2–24.4)1.00Reference 24.0–27.9 kg/m215.7 (13.9–17.6)29.5 (27.4–31.7)23.2 (21.7–24.7)0.550.45–0.66< 0.001 ≥ 28.0 kg/m212.1 (8.9–16.1)28.4 (24.4–32.8)21.5 (18.7–24.6)0.630.45–0.880.007 P for trend< 0.0010.113< 0.001SRH ≤ 6024.1 (22.7–25.4)38.4 (36.5–40.4)29.5 (28.4–30.7)1.00Reference > 6010.7 (9.4–12.2)19.3 (17.6–21.0)15.2 (14.1–16.4)0.430.39–0.48< 0.001 P for difference< 0.001< 0.001< 0.001WRII No17.4 (16.2–18.6)17.3 (15.2–19.5)17.4 (16.3–18.4)1.00Reference Yes25.0 (23.1–27.1)34.4 (32.8–36.1)31.0 (29.7–32.3)2.141.94–2.35< 0.001 P for difference< 0.001< 0.001< 0.001Type of pneumoconiosis Others8.1 (5.6–11.5)25.6 (22.0-29.6)18.6 (16.1–21.4)1.00Reference Silicosis18.6 (17.3–20.0)32.1 (29.9–34.4)23.2 (22.0-24.4)1.321.09–1.590.004 CWP23.5 (21.7–25.3)28.9 (27.0-30.8)26.2 (24.9–27.5)1.551.29–1.87< 0.001 P for difference< 0.0010.001< 0.001Stages of pneumoconiosis Stage I11.5 (10.1–13.1)28.4 (26.7–30.1)21.7 (20.5–22.9)1.00Reference Stage II19.7 (17.7–21.7)34.6 (31.7–37.7)25.5 (23.8–27.3)1.241.10–1.39< 0.001 Stage III28.8 (26.7–31.0)32.5 (28.4–37.0)29.6 (27.7–31.6)1.521.35–1.71< 0.001 Clinical cases (no classified stage)20.1 (17.2–23.2)18.5 (14.1–24.0)19.7 (17.2–22.4)0.880.74–1.060.172 P for trend< 0.0010.3440.054Pulmonary bullae or pneumothorax No16.1 (15.1–17.2)27.6 (26.2–29.0)21.3 (20.5–22.2)1.00Reference Yes46.5 (42.8–50.2)59.6 (53.9–65.2)50.4 (47.2–53.5)3.753.27–4.29< 0.001 P for difference< 0.001< 0.001< 0.001COPD Chronic obstructive pulmonary disease, OR Odds ratio, CI Confidence interval, BMI Body mass index, WRII Work-related injury insurance, SRH Self-Rated Health, CWP Coal workers’ pneumoconiosis Prevalence and Odds ratios of COPD associated with social determinants and clinical factors among pneumoconiosis patients in 27 provinces of China, stratified by urban and rural residence COPD Chronic obstructive pulmonary disease, OR Odds ratio, CI Confidence interval, BMI Body mass index, WRII Work-related injury insurance, SRH Self-Rated Health, CWP Coal workers’ pneumoconiosis Multivariate analysis of factors associated with COPD in patients with pneumoconiosis Prior to regression modeling, diagnostic checks confirmed that all variables had a tolerance > 0.1 and a variance inflation factor (VIF) < 10, indicating no concern for multicollinearity (Supplementary Table 3). In the model including all enrolled patients, significant factors associated with prevalent COPD among individuals with pneumoconiosis were identified as: increasing age, smoking index ≥ 200, BMI < 18.5 kg/m², residence in western or northeastern China, having WRII, comorbid cardiovascular diseases, CWP, advanced pneumoconiosis stage, cumulative dust exposure of 20–24 years, history of immune-enhancing drug use, history of hormone therapy for pneumoconiosis, and comorbid pulmonary bullae or pneumothorax (all P < 0.05). Notably, the association between WRII and higher COPD odds may reflect detection bias, as insured individuals typically undergo more systematic health monitoring, leading to higher diagnostic identification of COPD. Stratified analyses by urban/rural residence showed that the markedly higher COPD odds observed in the northeastern region was primarily confined to urban patients. The association of CWP with COPD was higher among urban patients than rural counterparts. Furthermore, case source, survey season, history of immune-enhancing drug use, and history of hormone therapy for pneumoconiosis were also significantly associated with COPD, with most of these associations differing between urban and rural populations (Table 2 and Supplementary Table 4). Table 2Multivariable-adjusted odds ratios (ORs) for factors associated with COPD among pneumoconiosis patients in 27 provinces of ChinaItemsAll pneumoconiosis casesLiving in rural areasLiving in urban areasOR (95%CI) P OR (95%CI) P OR (95%CI) P Age1.03 (1.03-1.04)<0.0011.03 (1.02-1.04)<0.0011.05 (1.04-1.06)<0.001Smoking index <1001.00 (Reference)1.00 (Reference)1.00 (Reference) 100-1991.08 (0.88-1.32)0.4491.27 (0.97-1.66)0.0800.88 (0.63-1.21)0.430 200-3991.60 (1.37-1.89)<0.0011.61 (1.30-2.00)<0.0011.67 (1.29-2.15)<0.001 >3991.77 (1.54-2.03)<0.0011.63 (1.35-1.97)<0.0011.88 (1.52-2.32)<0.001Alcohol intake Non-drinkers1.00 (Reference)1.00 (Reference)1.00 (Reference) Current drinkers0.79 (0.67-0.92)0.0020.64 (0.51-0.81)<0.0010.97 (0.78-1.21)0.806 Former drinkers1.03 (0.90-1.18)0.6551.26 (1.06-1.50)0.0090.94 (0.75-1.18)0.610BMI <18.5kg/m21.40 (1.14-1.70)0.0011.48 (1.17-1.88)0.0011.28 (0.86-1.89)0.222 18.5-23.9kg/m21.00 (Reference)1.00 (Reference)1.00 (Reference) 24.0-27.9kg/m20.90 (0.79-1.02)0.1010.86 (0.71-1.04)0.1140.90 (0.75-1.09)0.285 ≥28.0kg/m20.84 (0.67-1.04)0.1120.62 (0.42-0.92)0.0170.92 (0.69-1.22)0.559Region of China Eastern1.00 (Reference)1.00 (Reference)1.00 (Reference) Central0.79 (0.66-0.94)0.0091.97 (1.51-2.57)<0.0010.28 (0.20-0.37)<0.001 Western2.20 (1.88-2.58)<0.0013.15 (2.50-3.97)<0.0011.29 (0.99-1.67)0.058 Northeastern8.81 (7.20-10.79)<0.0011.88 (1.09-3.26)0.0249.70 (7.45-12.62)<0.001 With WRII1.37 (1.19-1.56)<0.0011.43 (1.20-1.71)<0.0011.35 (1.07-1.69)0.010 SRH>600.65 (0.57-0.74)<0.0010.61 (0.51-0.74)<0.0010.76 (0.63-0.91)0.003Types of pneumoconiosis Others1.00 (Reference)1.00 (Reference)1.00 (Reference) Silicosis1.24 (0.98-1.58)0.0731.53 (0.96-2.42)0.0721.27 (0.94-1.72)0.125 CWP2.18 (1.71-2.77)<0.0012.07 (1.31-3.28)0.0023.34 (2.43-4.60)<0.001Stages of pneumoconiosis Stage I1.00 (Reference)1.00 (Reference)1.00 (Reference) Stage II1.37 (1.19-1.58)<0.0011.45 (1.16-1.80)<0.0011.53 (1.24-1.88)<0.001 Stage III1.68 (1.43-1.98)<0.0011.98 (1.59-2.47)<0.0011.40 (1.04-1.89)0.028 Clinical cases (no classified stage)1.59 (1.26-2.00)<0.0011.65 (1.24-2.21)<0.0011.49 (0.94-2.37)0.092Pulmonary bullae or pneumothorax2.80 (2.36-3.33)<0.0012.56 (2.08-3.14)<0.0013.49 (2.50-4.86)<0.001COPD Chronic obstructive pulmonary disease, OR Odds ratio, CI Confidence interval, BMI Body mass index, WRII Work-related injury insurance, SRH Self-Rated Health, CWP Coal workers’ pneumoconiosis Multivariable-adjusted odds ratios (ORs) for factors associated with COPD among pneumoconiosis patients in 27 provinces of China COPD Chronic obstructive pulmonary disease, OR Odds ratio, CI Confidence interval, BMI Body mass index, WRII Work-related injury insurance, SRH Self-Rated Health, CWP Coal workers’ pneumoconiosis The nonlinear association between age and COPD The fully adjusted restricted cubic spline (RCS) regression model demonstrated a significant non-linear positive association between age and the prevalence of COPD among patients with pneumoconiosis (P for overall < 0.001; P for nonlinearity = 0.002). Using 55 years as the reference (OR = 1), the odds of COPD increased progressively with advancing age, with a more pronounced rise observed in later life (Fig. 1). Four knots were placed at the 5th, 35th, 65th, and 95th percentiles of the age distribution, corresponding to 42, 52, 61, and 80 years, respectively. Nonlinear association between age and COPD in patients with pneumoconiosis (fully adjusted model). OR Odds ratio, CI Confidence interval Subgroup analysis To further investigate the association between age and prevalent COPD, we conducted subgroup analyses and interaction tests. All subgroup models treated age as a continuous variable and were fully adjusted for covariates. As shown in Fig. 2 and Supplementary Fig. 1, significant effect modification was observed for residence, BMI, region of China, WRII, SRH, and types of pneumoconiosis (all P for interaction < 0.05). These results indicate that the strength of the association between age and COPD prevalence varied substantially across clinical and demographic subgroups. Subgroup and interaction analyses of the association between age and COPD in patients with pneumoconiosis (adjusted for other covariates). BMI Body mass index, WRII Work-related injury insurance, SRH Self-Rated Health, OR Odds ratio Stratified RCS analysis of the association between age and COPD To investigate heterogeneity in the association between age and COPD, we conducted stratified RCS analyses based on factors that showed significant interactions in subgroup analyses. Stratified analyses revealed substantial heterogeneity in age-related associations across subgroups. Among rural residents, individuals without WRII, those with SRH ≤ 60, individuals with a BMI of 18.5–23.9 kg/m², and those living in eastern regions, the odds of COPD increased steadily with advancing age (Fig. 3a and c; Supplementary Figs. 2–4). In contrast, among urban residents and individuals with WRII, the odds rose sharply after age 55 and gradually plateaued at older ages (Fig. 3b and d). Stratified analysis of age-related COPD by residence and work-related injury insurance(WRII) status in patients with pneumoconiosis (fully adjusted model). Rural residence (a), urban residence (b), without WRII (c) and with WRII d. OR Odds ratio, CI Confidence interval Pneumoconiosis subtype also significantly modified the association between age and COPD. Stratified restricted cubic spline analyses further revealed distinct age‑related patterns in COPD odds across subtypes: patients with silicosis (Supplementary Fig. 5a) exhibited a steeper age‑associated increase, whereas those with CWP (Supplementary Fig. 5b) showed a comparatively modest rise in COPD odds with age. Sensitivity analyses To validate the robustness of our primary findings, we conducted comprehensive sensitivity analyses that consistently supported the initial conclusions. To assess the stability of the reference age assumption, we re-fitted the restricted cubic spline (RCS) models using alternative reference ages of 45 and 50 years. These analyses consistently showed a significant nonlinear positive association between age and COPD (all P < 0.01; Supplementary Fig. 6). Importantly, the overall shape and steepness of the age-related odds curves remained largely consistent across different reference age specifications, indicating that the observed association was robust and not driven by the choice of the primary reference age. Second, to evaluate potential model dependency in the nonlinear association between age and COPD, we performed restricted cubic spline (RCS) analyses using both three‑knot and five‑knot specifications. Both configurations demonstrated a significant nonlinear positive association (both P < 0.05; Supplementary Fig. 7). Third, to minimize potential confounding, we constructed a balanced cohort using 1:1 nearest-neighbor propensity score matching, yielding 2,252 matched pairs of pneumoconiosis patients with and without COPD, with all standardized mean differences below 0.1 (Supplementary Table 5). Restricted cubic spline (RCS) analysis within the matched cohort consistently demonstrated a significant nonlinear positive association between age and COPD (P < 0.05). The overall curve shape and the age-related pattern were consistent with those observed in the primary analysis (Fig. 4). Propensity score matching (PSM) validation and age-related COPD after PSM in patients with pneumoconiosis (RCS model adjusted for covariates excluding PSM variables). a Propensity score density distribution before and after matching; (b) Non-linear association between age and COPD after PSM in patients with pneumoconiosis. OR Odds ratio, CI Confidence interval Finally, after excluding the clinical cases group, the nonlinear association between age and COPD remained robust (Supplementary Figs. 8, 9). Similarly, excluding participants from the Northeastern region yielded consistent results (Supplementary Fig. 10). In addition, removing pulmonary bullae and pneumothorax from the covariate set did not materially alter the RCS curve compared with the main model (Supplementary Fig. 11). Discussion In this nationwide study of 9,964 patients with pneumoconiosis, we delineate for the first time a nonlinear association between age and prevalent COPD that is substantially modified by sociodemographic factors, notably residence and work-related injury insurance status. These findings refine the understanding of COPD comorbidity in this high-risk occupational population and provide critical evidence for identifying high‑risk subgroups, which is of significant value for guiding public health planning, targeted population monitoring, and early intervention strategies. Although pneumoconiosis and COPD have traditionally been regarded as distinct clinical entities, accumulating evidence highlights considerable pathophysiological overlap. Both conditions involve chronic pulmonary inflammation, protease-antiprotease imbalance, and oxidative stress, which are core processes associated with progressive and irreversible airflow limitation [20, 21]. A meta-analysis reported a detection rate of 26.4% for comorbid COPD in patients with pneumoconiosis [22]. In our study, the prevalence of COPD among pneumoconiosis patients was 24.1%, which is consistent with this pooled estimate and significantly higher than the 13.6% reported in the general Chinese population aged 40 years and above [23]. This finding supports the notion that long-term occupational dust exposure represents an independent associated factor for COPD [24]. We further identified a graded association between pneumoconiosis severity and COPD prevalence [7], reinforcing the concept that dust-induced structural lung damage and sustained inflammatory responses are linked to the presence of COPD [25]. This study demonstrates that age is independently associated with prevalent COPD in patients with pneumoconiosis, consistent with findings from most epidemiological studies [6, 9, 26]. With advancing age, physiological changes including reduced pulmonary elasticity and impaired reparative capacity, combined with the inherent pulmonary fibrosis and inflammatory damage of pneumoconiosis, may exert a synergistic effect that is associated with the presence of COPD [27, 28]. Notably, RCS analysis demonstrated a significant nonlinear positive association, with the age–odds curve showing a pronounced acceleration around 55 years of age. This pattern remained robust in sensitivity analyses, persisting when the reference age was varied and after propensity score matching to control for baseline confounders. The steeper increase in COPD odds at this stage likely reflects the cumulative synergistic effects of prolonged occupational dust exposure and age-related physiological decline, which may progressively erode pulmonary reserve and increase susceptibility to COPD. Importantly, the marked rise in COPD odds among Chinese patients with pneumoconiosis occurs earlier than the age at which COPD prevalence typically increases rapidly in the general populations of high sociodemographic index (SDI) countries (approximately 60 years) [12, 29]. This temporal shift suggests that heavy occupational exposure may contribute to an earlier onset of elevated COPD odds, underscoring the need for earlier respiratory surveillance in this vulnerable workforce. Critically, the association between age and COPD was not uniform but was substantially modified by sociodemographic and clinical characteristics. Clear differences were observed between urban and rural residents. Among urban residents, the odds of COPD increased sharply after middle age and then gradually plateaued in later life. In contrast, rural residents exhibited a continuous increase in COPD odds with advancing age. This disparity remained robust in sensitivity analyses following propensity score matching. The difference may be partly explained by uneven medical resource allocation and health screening coverage. Urban residents generally have better access to routine health examinations, which facilitates earlier detection of COPD in middle‑to‑older age. By contrast, rural residents often face limited screening opportunities, resulting in a sustained rise in COPD odds across older adulthood [30]. A similarly complex pattern was observed with respect to WRII status. Although insured individuals had a higher prevalence of COPD (31.0% vs. 17.4%) and WRII coverage emerged as an independent associated factor in multivariable analysis (OR = 1.37), this finding likely reflects detection bias arising from more systematic health surveillance. Stratified restricted cubic spline analyses further showed that insured individuals experienced a sharp increase in COPD odds after middle age, followed by a gradual plateau in later life, whereas uninsured individuals exhibited a steady increase in COPD odds with advancing age. This divergent age‑related pattern was consistently replicated in the matched cohort. Improved healthcare access and regular occupational health examinations among individuals with WRII likely facilitate earlier detection of COPD in middle‑to‑older age, whereas the absence of routine screening among uninsured individuals contributes to a sustained increase in COPD odds across older adulthood [31, 32]. A notable finding of this study is the markedly elevated odds of COPD among pneumoconiosis patients in Northeastern China (OR = 8.81), particularly in urban areas. This regional concentration likely reflects the distinctive industrial and environmental context of the region. As China’s traditional heavy‑industrial heartland, Northeastern China is characterized by large state‑owned industrial complexes, where urban workers have historically experienced higher cumulative occupational dust exposure and longer exposure durations than workers in rural mining sectors in other regions. In addition, the region’s harsh winter climate and long-standing reliance on coal‑fired centralized heating systems contribute to a “double burden” of indoor and outdoor air pollution, which may synergistically accelerate COPD progression among individuals with pneumoconiosis. Importantly, sensitivity analyses excluding the Northeastern cohort produced consistent results regarding the non-linear association between age and COPD, supporting the robustness and generalizability of the study’s primary conclusions. Pneumoconiosis subtype also emerged as an important modifier of the association with COPD. The prevalence of COPD differed significantly between subtypes, with a higher prevalence observed in CWP than in silicosis. Multivariable regression analysis further confirmed that CWP was independently associated with increased odds of COPD. Stratified restricted cubic spline analyses showed that the age‑related increase in COPD odds was more pronounced among individuals with silicosis than among those with CWP. These subtype‑specific differences may reflect distinct exposure patterns and underlying pathological mechanisms. Silicosis is primarily caused by long‑term exposure to crystalline silica dust, which accelerates lung function decline with advancing age and contributes to a steeper age‑related increase in COPD odds. In contrast, CWP results mainly from the deposition of mixed coal dust; its pathological progression appears relatively stable across age, leading to a less pronounced age‑related trend [33]. Together, these findings highlight the importance of incorporating pneumoconiosis subtype into the stratification of COPD odds and targeted clinical monitoring strategies. This study demonstrates a significant nonlinear association between age and prevalent COPD among patients with pneumoconiosis, an association that is substantially modified by sociodemographic and clinical factors. The age–odds curve shows the steepest increase at approximately 55 years of age. Notably, place of residence and WRII status emerged as key modifiers of this relationship, underscoring the influence of structural health inequities. Overall, these individuals should be prioritized for enhanced pulmonary function monitoring and targeted early interventions. Addressing sociodemographically driven disparities is critical to reducing the disproportionate burden of COPD in this vulnerable occupational population. We acknowledge several important limitations. First, the cross-sectional design precludes causal inference, and survival (selection) bias cannot be excluded. Patients with pneumoconiosis who develop early-onset or severe COPD may experience higher mortality and therefore be underrepresented in the surveyed population. This could shift the observed age–COPD association toward older ages and lead to an underestimation of the true age-related increase, particularly among socioeconomically disadvantaged subgroups where early mortality may be more pronounced. As a result, the actual disparities in COPD odds between groups may be larger than those observed in this study. Second, several variables, including self-rated health and cumulative dust exposure duration, were self-reported and therefore subject to potential recall bias. Third, because spirometry was conducted across multiple centers, residual inter-site measurement variability may have persisted despite standardized procedures and routine calibration or calibration verification. This variability may have affected COPD classification among participants with FEV1/FVC values near the diagnostic threshold. Moreover, although dust exposure duration was obtained through questionnaire data, information on exposure intensity or cumulative exposure dose was not available, which may reduce the precision of the exposure–outcome analyses. Future prospective cohort studies incorporating objective exposure measurements including dust exposure intensity and longitudinal lung function monitoring across multinational populations are needed to further elucidate these associations and clarify potential causal mechanisms. Supplementary Information Supplementary Material 1. Acknowledgements The authors extend their gratitude to all study participants, the survey teams across the 27 provinces, and the project development and management teams for their invaluable contributions. Authors’ contributions Conception and design: H W, K L, Y R. Analysis and interpretation of data: K L, L Z, X L. Drafting the article or revising it critically for important intellectual content: K L, H W. Final approval of the version to be submitted for revision: H W, K L, Y R. All authors had full access to all the study data and accepted responsibility for submitting this work. Funding National Major Science and Technology Project of China (No. 2024ZD0528901); National Natural Science Foundation of China (No. 82574060). Data availability The original individual participant data cannot be shared publicly due to containing sensitive personal information. De-identified summary data are available from the corresponding author upon reasonable request and with ethics committee approval. Declarations Ethics approval and consent to participate This study was approved by the Ethics Committee of the National Institute of Occupational Health and Poison Control, Chinese Center for Disease Control and Prevention (Approval No. 201720). Written informed consent was obtained from all participants prior to data collection. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The original individual participant data cannot be shared publicly due to containing sensitive personal information. De-identified summary data are available from the corresponding author upon reasonable request and with ethics committee approval.