Abstract
BACKGROUND: Chronic obstructive pulmonary disease (COPD) poses a major global health burden with high morbidity. Relative fat mass (RFM), as a novel body fat measurement indicator, can reflect the distribution of body fat. This study aims to elucidate its associations with COPD prevalence and severity in two cohorts to enhance prevention and treatment strategies. METHODS: We retrospectively investigated the medical records of 166 patients with COPD and the data of 2654 subjects from two cohorts. To explore the relative importance of factors in COPD prevalence and severity, we built an extreme gradient boosting (XGBoost) machine-learning model. Logistic regression models were used to assess the relationship between COPD and RFM, with subgroup analysis to clarify the difference across diverse subgroups. Furthermore, restricted cubic spline (RCS) curves were used to explore the exposure-response relationship. RESULTS: Multivariate logistic regression analysis revealed a significant positive association between RFM and COPD prevalence (OR = 1.043, 95% CI: 1.004-1.083, p = 0.030) and a negative association with COPD severity (OR = 0.892, 95% CI: 0.813-0.978, p = 0.015). According to the RCS curves, there was no nonlinear association between RFM and COPD prevalence or severity (p for nonlinear = 0.703, p for nonlinear = 0.348). CONCLUSION: RFM was positively associated with the prevalence of COPD but inversely associated with its severity. Specifically, RFM predicted COPD prevalence more accurately in individuals aged 40-60 and smokers, while it predicted COPD severity more effectively in those aged ≥ 60.