Skip to main content
NUTRITION 2023 Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Back

Session: Big Data and Omics in Bioactives Research (Oral 15)

(OR15-06-23) Untargeted Metabolomics of Human Plasma and Foods Enable Identification of Putative Health-Driven Biomarkers

Monday, July 24, 2023
9:00 AM – 9:12 AM ET
Location:  Commonwealth
Disclosure(s):
Audrey E. Hendricks, PhD: No relevant financial relationship(s) with ineligible companies to disclose.
Emily B. Hill, PhD, RDN: No relevant financial relationship(s) with ineligible companies to disclose.
Sarah J. Borengasser, PhD: No relevant financial relationship(s) with ineligible companies to disclose.
Lauren E. O'Connor, PhD, MPH: No relevant financial relationship(s) with ineligible companies to disclose.
Nancy F. Krebs, MD, MS: Nurture,Inc. Danone: Advisory Committee/Board Member (Terminated)
Objectives: To identify plasma compounds associated with change in cardiovascular (CV) health indicators and relate these back to food.

Methods: Untargeted metabolomics data were generated from plasma samples collected within a randomized, cross-over Mediterranean diet (MED) trial (N=41) with multiple CV indicators measured pre- and post-intervention. In depth quality control (QC) was performed including batch and order correction. Using all compounds that passed QC, we estimated variability in the change of CV indicators explained by compounds identified in the lipid or aqueous fractions of prepared plasma. We used a linear mixed-effects model to assess association between change in CV indicators and change in individual compounds. We linked these compounds back to foods using a complementary dataset of untargeted metabolomics of 100 foods consumed in the MED trial. A false discovery rate (FDR) of 0.05 was used to control for multiple testing and assess statistical significance.


Results: After QC, 2296 lipid and 2356 aqueous plasma compounds remained. While lipid compounds explained a significant and high (~75 to 83%) proportion of variance in lipid related indicators (e.g., total cholesterol, LDL), they did not explain a significant proportion of the variance in non-lipid related indicators (~18-26%; e.g., C-reactive protein, blood pressures). Aqueous compounds explained a small and nonsignificant proportion of variance in all health indicators. At FDR < 0.05, 489 lipid and 5 aqueous compounds changed after the diet intervention and were associated with change in at least one CV indicator. Of these, 136 lipid compounds were annotated and matched to compounds in 40 foods. For instance, Triricinolein and Inositol-P-ceramide increased over the MED trial (FDR=2.34e-5 and 7.13e-4), were associated with a decrease in total cholesterol (FDR=4.42e-2 and 1.33e-3) and were found in blueberries.

Conclusions: Changes in compounds detected by untargeted metabolomics of the lipid fraction, but not the aqueous fraction, of human plasma were related to changes in CV health indicators in a controlled, Mediterranean diet intervention. Several compounds can be traced back to specific foods indicating potential utility as food-driven biomarkers of health.

Funding Sources: NIH/NIDDK; Beef Checkoff & Pork Checkoff (feeding study)