Identification of serum N,N-dimethylglycine as a potential biomarker for prenatal diagnosis of congenital heart disease using HNMR and UPLC-MS/MS metabonomics.

Congenital heart disease (CHD) poses significant clinical challenges due to limitations in early prenatal diagnosis. This study aimed to identify serum metabolic biomarkers for CHD using a combined metabolomics approach. Serum samples from 55 pregnant women carrying fetuses with confirmed CHD (CHDP group) and 49 healthy controls (ZCP group) were analyzed via non-targeted HNMR metabolomics, revealing distinct metabolic profiles. Six choline pathway metabolites were further quantified by ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). Among these, N,N-dimethylglycine (DMG) exhibited the most significant reduction in the CHDP group (1.07 ± 0.03 vs. 1.55 ± 0.04 µg/mL, p < 0.001), with an area under the ROC curve (AUROC) of 0.883 in the discovery cohort. Validation in an independent cohort (58 CHDP vs. 62 ZCP) confirmed DMG's diagnostic potential (AUROC = 0.818). While betaine-homocysteine methyltransferase 2 (BHMT2) activity showed no intergroup differences, DMG's consistent performance highlights its utility as a non-invasive biomarker. This study underscores the clinical value of metabonomics in prenatal CHD screening and establishes DMG as a promising diagnostic marker, potentially improving early detection and perinatal management.