Longitudinal reciprocal association between rheumatoid arthritis and chronic obstructive pulmonary disease and mediation of inflammation.

Abstract

Objectives: To elucidate the longitudinal reciprocal association between RA and chronic obstructive pulmonary disease (COPD) and the mediating role of systemic inflammation in the association.

Methods: A total of 403 045 participants from UK Biobank were enrolled in this study. A cross-lagged panel model was used to investigate the longitudinal reciprocal association between RA and COPD. Cox proportional hazards regression and logistic regression models were also conducted to examine the association between baseline RA and COPD during follow-up, and vice versa. Causal mediation analysis was then performed to explore the mediating roles of 160 systemic inflammatory biomarkers in the bidirectional association.

Results: At baseline, 4755 (1.2%) and 6989 (1.7%) individuals were diagnosed with RA and COPD, respectively. After adjusting for the covariates, the result of a cross-lagged panel model revealed a bidirectional association between RA and COPD (β = 0.018, P < 0.001 for the RA→COPD path; β = 0.010, P < 0.001 for the COPD→RA path). In the non-COPD population, the risk of future COPD was increased in RA patients [Cox model: hazard ratio (HR) 1.65 (95% CI 1.50, 1.83); logistic model: odds ratio (OR) 1.85 (95% CI 1.66, 2.07)]. In the non-RA population, baseline COPD was associated with a higher risk of RA during follow-up [Cox model: HR 1.67 (95% CI 1.44, 1.92); logistic model: OR 1.70 (95% CI 1.47, 1.97)]. Five inflammatory factors mediated the RA→COPD path and CRP mediated the COPD→RA path (false discovery rate < 0.05).

Conclusions: A significant bidirectional association exists between RA and COPD and it is partially mediated by systemic inflammation.