A cross-omics integrative study of metabolic signatures of chronic obstructive pulmonary disease

• Main Authors: Ivana Prokić, Lies Lahousse, Maaike de Vries, Jun Liu, Marita Kalaoja, Judith M. Vonk, Diana A. van der Plaat, Cleo C. van Diemen, Ashley van der Spek, Alexandra Zhernakova, Jingyuan Fu, Mohsen Ghanbari, Mika Ala-Korpela, Johannes Kettunen, Aki S. Havulinna, Markus Perola, Veikko Salomaa, Lars Lind, Johan Ärnlöv, Bruno H. C. Stricker, Guy G. Brusselle, H. Marike Boezen, Cornelia M. van Duijn, Najaf Amin
• Format: Article
• Language: English
• Published: BMC 2020-07-01
• Series: BMC Pulmonary Medicine
• Subjects: COPD; Metabolomics; Mendelian randomization; Glycoprotein acetyls; Biomarkers
• Online Access: http://link.springer.com/article/10.1186/s12890-020-01222-7

Abstract Background Chronic obstructive pulmonary disease (COPD) is a common lung disorder characterized by persistent and progressive airflow limitation as well as systemic changes. Metabolic changes in blood may help detect COPD in an earlier stage and predict prognosis. Methods We conducted a comprehensive study of circulating metabolites, measured by proton Nuclear Magnetic Resonance Spectroscopy, in relation with COPD and lung function. The discovery sample consisted of 5557 individuals from two large population-based studies in the Netherlands, the Rotterdam Study and the Erasmus Rucphen Family study. Significant findings were replicated in 12,205 individuals from the Lifelines-DEEP study, FINRISK and the Prospective Investigation of the Vasculature in Uppsala Seniors (PIVUS) studies. For replicated metabolites further investigation of causality was performed, utilizing genetics in the Mendelian randomization approach. Results There were 602 cases of COPD and 4955 controls used in the discovery meta-analysis. Our logistic regression results showed that higher levels of plasma Glycoprotein acetyls (GlycA) are significantly associated with COPD (OR = 1.16, P = 5.6 × 10− 4 in the discovery and OR = 1.30, P = 1.8 × 10− 6 in the replication sample). A bi-directional two-sample Mendelian randomization analysis suggested that circulating blood GlycA is not causally related to COPD, but that COPD causally increases GlycA levels. Using the prospective data of the same sample of Rotterdam Study in Cox-regression, we show that the circulating GlycA level is a predictive biomarker of COPD incidence (HR = 1.99, 95%CI 1.52–2.60, comparing those in the highest and lowest quartile of GlycA) but is not significantly associated with mortality in COPD patients (HR = 1.07, 95%CI 0.94–1.20). Conclusions Our study shows that circulating blood GlycA is a biomarker of early COPD pathology.