Similarities of metabolomic disturbances in prematurity-associated obstructive lung disease to chronic obstructive pulmonary disease.

Prematurity-associated lung disease (PLD) is a long-term consequence of preterm-birth. Since the underlying mechanisms of PLD remain poorly characterised, we compared the urinary metabolome between recently described spirometry phenotypes of PLD. Preterm- and term-born children aged 7-12 years, from the Respiratory Health Outcomes in Neonates (RHiNO) cohort, underwent spirometry and urine collection. The urinary metabolome was analysed by gas chromatography time-of-flight mass spectrometry. Preterm-born children were classified into phenotypes of prematurity-associated obstructive lung disease (POLD, Forced expiratory volume in 1 s (FEV) < lower limit of normal (LLN), FEV/Forced Vital Capacity (FVC) < LLN), prematurity-associated preserved ratio impaired spirometry (pPRISm, FEV < LLN, FEV/FVC ≥ LLN) and Preterm/Term controls (FEV ≥ LLN). Metabolite set enrichment analysis was used to link significantly altered metabolites between the groups with metabolic pathways. Univariable and multivariable linear regression models examined associations between early and current life factors and significantly altered metabolites of interest. Urine from 197 preterm- and 94 term-born children was analysed. 23 and 25 were classified into POLD and pPRISm groups respectively. Of 242 identified metabolites, 49 metabolites were significantly altered in the POLD group compared with Preterm controls. Decreased capric acid (log fold change - 0.23; p = 0.003), caprylic acid (- 0.18; 0.003) and ceratinic acid (- 0.64; 0.014) in the POLD group, when compared to preterm controls, were linked with reduced β-oxidation of very long chain fatty acids (p = 0.004). Reduced alanine (log fold change - 0.21; p = 0.046), glutamic acid (- 0.24; 0.023), and pyroglutamic acid (- 0.17; 0.035) were linked with decreased glutathione metabolism (p = 0.008). These metabolites remained significantly associated with POLD in multivariable models adjusting for early/current life factors. The pPRISm urinary metabolome was minimally changed when compared with preterm-born controls. When compared to term-born subjects, alterations in tryptophan metabolism were implicated (p = 0.01). The urinary metabolome in POLD showed significantly altered β-oxidation of fatty acids and glutathione metabolism, implying alterations in cellular metabolism and oxidative stress. Similar findings have been noted in adults with chronic obstructive pulmonary disease. Given the similarity of findings between the POLD group and those reported for COPD, the POLD group should be considered at future risk of developing COPD.