Chronic obstructive pulmonary disease (COPD) is a leading cause of death globally, with occupational exposures, such as cotton dust in textile workers, significantly contributing to its development. While cigarette smoking is a primary risk factor, a substantial portion of COPD cases occur in non-smokers, highlighting the role of other factors. Occupational exposure to cotton dust, particularly the gram-negative bacterial endotoxin it often contains, has been linked to both acute and chronic respiratory issues, including accelerated decline in FEV₁. Previous research on the Shanghai Textile Workers Cohort demonstrated associations between cotton dust exposure and endotoxin with decreased FEV₁. This study aimed to identify novel protein biomarkers associated with long-term FEV₁ decline in this cohort, leveraging the advancements in large-scale proteomic technologies. The identification of such biomarkers could provide valuable insights into disease pathogenesis, facilitate early risk identification, and potentially inform the development of novel therapeutic targets for early intervention.

Several studies have examined specific protein biomarker panels in relation to lung function, but these were limited in scope. Recent developments in large-scale proteomic technologies allow for the simultaneous measurement of numerous proteins. A recent study across six cohorts identified and validated proteins associated with FEV1 decline, but research specifically focusing on the mechanisms underlying FEV1 decline in occupationally exposed populations remains limited. The Shanghai Textile Worker Study, initiated in 1981, is unique in its extensive longitudinal follow-up (35 years with eight lung function assessments), providing a robust dataset for exploring protein biomarkers related to occupational lung disease.

The Shanghai Textile Workers Cohort, established in 1981, included cotton and silk workers (exposed and unexposed controls, respectively). Lung function was assessed every five years for 35 years. In the 2016 survey, serum samples from 453 workers were collected for quantitative serum proteomics using data-independent acquisition mass spectrometry. Four models were used to analyze the association between FEV₁ and protein levels: cluster-based model, restricted cubic spline (RCS) model, latent class mixed model (LCMM), and mixed model for repeated measurements (MMRM). The results were combined using an aggregated Cauchy association test (ACAT). The UK Biobank (UKB) proteomic data was used for external exploration, examining associations between FEV₁ decline and proteins. Protein-protein interaction (PPI) networks were analyzed using STRING. Finally, two-sample Mendelian randomization (MR) was performed to assess causal relationships between identified proteins and FEV₁ decline.

After quality control and imputation of missing data, 15 proteins showed significant associations with long-term FEV₁ decline (FDR-q_ACAT < 0.05) in the Shanghai Textile Workers Cohort. These included two hemoglobin subunits (HBB, HBA2) and four immunoglobulin subunits (IGKV3-7, IgH). In the UK Biobank, five proteins (ART3, RAB6A, LRRN1, ANGPTL7, and BSG) were significantly associated with the rate of FEV₁ decline. PPI network analysis revealed connections between proteins identified in both cohorts, notably links between HBB and proteins like BSG, RAB6A, and LRRN1. Two-sample MR analysis indicated a bidirectional association between HBB and FEV₁, with a negative correlation observed from FEV₁ to HBB and a positive correlation in the reverse direction. Analyses of the immunoglobulin family proteins showed significant negative correlations between multiple immunoglobulins and FEV₁.

This study identified several novel proteins associated with long-term FEV₁ decline in textile workers. The identified proteins, including hemoglobin subunits and immunoglobulins, suggest potential mechanisms involving oxygen transport, inflammation, and immune responses. The multi-model approach and use of UK Biobank data and MR strengthened the findings. The bidirectional association of HBB and FEV₁ highlights the complex interplay between oxygen transport and lung function decline. The association of immunoglobulins suggests a possible role for immune response in long-term lung damage. While the study primarily focused on identifying biomarkers, the findings provide insights into the pathophysiology of lung function decline in occupational settings and may lead to earlier identification of at-risk individuals and better treatment strategies.

This large-scale proteomic study identified multiple novel proteins associated with long-term FEV₁ decline in a textile worker cohort. Hemoglobin subunits and immunoglobulin subunits emerged as potential biomarkers. The findings were supported by UK Biobank data and Mendelian randomization, suggesting potential causal relationships. Future research should investigate the underlying biological mechanisms and pathways involved, potentially leading to early intervention strategies for occupational lung disease.

The study's cross-sectional nature at the final time point prevents definitive causal inference. The healthy worker survivor effect may have influenced the results. The cohort included only Han Chinese individuals, limiting the generalizability of the findings to other populations. Although UK Biobank data and MR analysis provided some external validation, the heterogeneity of populations between the Shanghai Textile Worker Study and the UK Biobank remains a limitation. Finally, the use of serum proteomics, rather than respiratory tract proteomics, may have missed some relevant biomarkers.