Characterizing the metabolome of children with growth hormone deficiency

OBJECTIVES: Growth hormone deficiency (GHD) diagnosis requires inadequate GH responses to two provocative tests, which are time-consuming and may cause side effects. Recent advancements in serum metabolomics offer potential novel biomarkers for medical conditions. This study investigated serum metabolomics in children with GHD to explore new diagnostic approaches and identify altered biological pathways. METHODS: We conducted a prospective study of 68 children (aged 3-18 years) undergoing growth hormone stimulation tests (GHST). Children with genetic syndromes, systemic illnesses, or end-stage renal disease were excluded. Untargeted metabolomics analysis using liquid chromatography-mass spectrometry (LC-MS) identified 951 circulating metabolites (280 polar and 671 lipids). From the 68 children evaluated, 25 children were diagnosed with GHD, and 41 children served as controls. Two children exhibited a suboptimal GH peak during the first GHST but did not undergo a second confirmatory test. RESULTS: Significant differences were observed in 7 polar metabolites and 50 lipids between groups, but only phosphatidylserine (PS) (40:3) remained significant after false discovery rate correction. Cluster analysis revealed two lipid clusters significantly associated with GHD. Greater separation in metabolomic profiles was observed when a lower GH threshold was applied for diagnosis. CONCLUSIONS: This study provides proof of concept for a unique lipidomics profile in children with GHD, highlighting its potential as a diagnostic tool. Larger-scale studies are required to validate these findings.