Scientific Reports (Aug 2025)

Quantitative 1H-NMR spectroscopy identifies metabolites and lipoprotein subclasses associated with intermediate phenotypes of chronic diseases in the Japanese Nagahama Study

  • Huiting Ou,
  • Shuji Kawaguchi,
  • François Brial,
  • Kazuhiro Sonomura,
  • Takahisa Kawaguchi,
  • Andrée E. Gravel,
  • Sanjoy Kumar Das,
  • Daniel Auld,
  • Yasuharu Tabara,
  • Bertrand J. Jean-Claude,
  • Mark Lathrop,
  • Fumihiko Matsuda,
  • Dominique Gauguier

DOI
https://doi.org/10.1038/s41598-025-12305-y
Journal volume & issue
Vol. 15, no. 1
pp. 1 – 14

Abstract

Read online

Abstract Metabolomics is a powerful molecular phenotyping technology which can be used in population studies to identify metabolites underlying disease conditions. To identify plasma biomarkers potentially predicting chronic diseases we applied 1H nuclear magnetic resonance (NMR) metabolomics using a 600 MHz spectrometer fitted with an In Vitro Diagnostics Research (IVDr) platform to test associations between 18 known metabolites and 111 lipoprotein constituents that could be quantified and passed our quality control procedure and 944 phenotypes determined in 302 healthy participants of the Japanese Nagahama Study. We identified 907 statistically significant associations (p < 4.11 × 10–7) between 34 phenotypes and at least one metabolite or lipoprotein. Eight metabolites and 109 lipoprotein (sub)classes showed evidence of associations with phenotypes predominantly related to lipid and cholesterol metabolism, liver function, fatness and hematology. We confirmed previously reported associations between plasma trimethylamine-N-oxide (TMAO) and cholesterol, and between the branched-chain amino acids leucine and valine and body mass index (BMI). BMI and fatness were positively associated with components of plasma LDL-4 and VLDL-1 and the ratios of apolipoproteins A1 to B100 and LDL to HDL cholesterol, whereas they were inversely associated with HDL-1 constituents. HDL-1 and LDL-4 subclasses systematically follow the patterns of association of HDL and LDL, respectively, and we propose that these can be examined to improve cardiometabolic risk evaluation. Results from our study exemplify the power of quantitative NMR-based metabolome profiling applied to even relatively small cohorts of healthy individuals extensively characterized for multiple phenotypes underlying unrelated clinical conditions to identify potentially disease-predicting metabolite biomarkers.

Keywords