Background: Gestational diabetes mellitus (GDM) poses a risk of short-term and long-term complications for both mother and fetus. However, there is a lack of consensus on the screening approach and pathophysiology of GDM.
Methods: Women were screened at 24 to 28 weeks gestation using the one-step screening approach and serum samples were collected for metabolomics based on 1H NMR spectroscopy. A random forest classifier was developed to evaluate its diagnostic efficacy on GDM.
Results: Serum metabolic fingerprints of women with GDM differed significantly from those with normoglycemic. Of the 59 differential metabolites identified, 25 were well-known risk metabolites associated with type 2 diabetes or cardiovascular diseases, such as branched-chain amino acids and trimethylamine N-oxide. In addition, most of the differential metabolites were microbial metabolites or could be metabolized by gut microbes. The correlation between serum metabolites and maternal 75 g OGTT glucose values supported the establishment of a random forest classifier, which selected 21 metabolites to predict GDM with an AUC of 0.988.
Conclusions: Metabolic disturbances in the host and gut microbiota may be a persistent contributor to the risk of developing type 2 diabetes or cardiovascular diseases in GDM. Targeting microbiota is one intervention that needs to be considered.
Keywords: Gestational diabetes mellitus; Gut microbiota; Metabolism; Metabolomics; Type 2 diabetes.
Copyright © 2023 The Authors. Published by Elsevier B.V. All rights reserved.