Background and aim: Gut microbiota may regulate metabolism but is incompletely characterized in pregnancy. Our objective was to investigate the relations using omics techniques.

Methods and results: In a cross-sectional setting, fecal and serum samples of 361 healthy pregnant women with overweight or obesity were analyzed with a combinatorial approach of metagenomics and targeted NMR-based metabolomics, with statistical and machine learning techniques to identify and analyze the extent to which the gut microbiota composition and predicted functions would be reflected in the serum metabolome. We identified five biclusters, each of which consisted of a set of gut microbial species and serum metabolites with correlated abundance profiles. Two of the biclusters included metabolites that have been linked to the cardiovascular health; one was linked with factors known to increase the risk i.e., various sizes of lipoprotein subclasses (VLDL and LDL), subclasses of relative lipoprotein lipid concentrations (VLDL, IDL, and LDL), apolipoprotein B, and an inflammation marker, glycoprotein acetylation. These metabolites were associated with abundances of species such as, Enterocloster bolteae and Ruminococcus gnavus. The second bicluster included metabolites linked with a reduced cardiovascular risk, such as different sizes of HDL (high-density lipoprotein), subclasses for relative lipoprotein lipid concentrations and mean diameter for HDL particles, and fatty acid ratios. These metabolites were associated with abundances of species, such as Bacteroides cellulosilyticus and Alistipes finegoldii. We did not observe any biclusters between predicted pathways and serum metabolites.

Conclusion: Overall, we identified five biclusters of co-abundant gut bacteria and serum metabolites, of which two were linked to pro-atherogenic and anti-atherogenic properties.

Trial registration: www.ClinicalTrials. Gov: NCT01922791.