Cystic fibrosis (CF) is a multi‐system disease affecting multiple organs and cells besides the respiratory system. Metabolomic profiling allows simultaneous detection of biochemicals originating from cells, organs, or exogenous origin that may be valuable for monitoring of disease severity or in diagnosis.

We hypothesized that metabolomics using serum from children would differentiate CF from non‐CF lung disease subjects and would provide insight into metabolism in CF.

Serum collected from children with CF (n = 31) and 31 age and gender matched children with other lung diseases was used for metabolomic profiling by gas‐ and liquid‐chromatography. Relative concentration of metabolites was compared between the groups using partial least square discriminant analyses (PLS‐DA) and linear modeling.

A clear separation of the two groups was seen in PLS‐DA. Linear model found that among the 459 detected metabolites 92 differed between CF and non‐CF. These included known biochemicals in lipid metabolism, oxidants, and markers consistent with abnormalities in bile acid processing. Bacterial metabolites were identified and differed between the groups indicating intestinal dysbiosis in CF. As a novel finding several pathways were markedly different in CF, which jointly point towards decreased activity in the β‐oxidation of fatty acids. These pathways include low ketone bodies, low medium chain carnitines, elevated di‐carboxylic acids and decreased 2‐hydroxybutyrate from amino acid metabolism in CF compared to non‐CF.

Serum metabolomics discriminated CF from non‐CF and show altered cellular energy metabolism in CF potentially reflecting mitochondrial dysfunction. Future studies are indicated to examine their relation to the underlying CF defect and their use as biomarkers for disease severity or for cystic fibrosis transmembrane regulator (CFTR) function in an era of CFTR modifying drugs. Pediatr Pulmonol. 2014; 49:463–472. © 2013 Wiley Periodicals, Inc.