Although necrotising enterocolitis (NEC) carries significant morbidity/mortality in preterm babies, it is poorly understood and its identification is hindered by the lack of a single diagnostic test. Metabolomics offers innovative opportunities for biomarker discovery and improving pathogenetic insight.

We implemented a novel pilot study to determine if metabolomics could identify preterm babies at risk of NEC. To aid data interpretation, we also assessed the impact of gestational age and feeding on the metabolome.

We collected serum samples from 12 pre-terms (<30 weeks’ gestation) and 8 controls (≥37 weeks’). Two serum samples were taken (before 7 days of age and when fully enterally fed) from each infant. A third sample was obtained only on developing NEC. Samples underwent gas/liquid chromatography-mass spectrometry (GC-MS/LC-MS) analysis before undergoing univariate, multivariate and pathway analysis.

No metabolites could reliably identify infants at risk of developing NEC. Although the metabolome appears to alter in response to NEC, sample size prevented detailed analysis of this and of feeding. 3983 of 10419 and 54 of 166 metabolites, detected in LC-MS and GC-MS respectively, significantly differed between preterm and term babies, particularly fatty acids (Kruskal-Wallis p-value < 0.05). Furthermore, 12 significant metabolites detected by GC-MS were mapped to one network linked to carbohydrate/lipid metabolism (p = 1 × 10^–30).

Whilst we could not identify predictive biomarkers for NEC we observed significant differences in the metabolomes of preterms and terms. Provisional pathway analysis of these differences suggests altered insulin action in preterm infants. This pilot study indicates that metabolomics may be used to study glucose control in preterm babies.