Abstract
BACKGROUND: Respiratory failure (RF) is the leading cause of death in chronic obstructive pulmonary disease (COPD), yet reliable biomarkers for early risk stratification remain unclear. Circulating C-reactive protein (CRP) reflects systemic inflammation, but its prognostic value for incident RF in COPD is controversial. METHODS: A total of 38 933 patients from the UK Biobank with the ratio of Forced Expiratory Volume in 1 second to Forced Vital Capacity (FEV(1)/FVC) < 0.70 but without RF at baseline were included, and a maximum of 17.87 years of follow-up was conducted. Participants were divided into five subgroups based on serum CRP concentration. Kaplan-Meier survival analysis was utilised to assess the correlation between CRP stratification, incident RF, all-cause mortality, and COPD-induced mortality. The dose-response relationship between CRP concentration and incident RF was investigated using Cox proportional hazards regression. RESULTS: Kaplan-Meier curves showed statistically significant differences in RF across all subgroups throughout the entire follow-up period. Additionally, significant differences were observed between groups concerning all-cause mortality and COPD-induced mortality as well. The Cox proportional hazards model demonstrated a clear dose-response relationship between CRP concentration and RF, even after adjustment for several clinical covariates and systemic inflammation index. CONCLUSIONS: Serum CRP concentration may forecast a high risk of incident RF in patients with COPD, indicating further research on the threshold.