Biomarkers of Individual Foods, and Separation of Diets Using Untargeted LC-MS-based Plasma Metabolomics in a Randomized Controlled Trial

Abstract

Scope: Self-reported dietary intake does not represent an objective unbiased assessment. The effect of the new Nordic diet (NND) versus average Danish diet (ADD) on plasma metabolic profiles is investigated to identify biomarkers of compliance and metabolic effects.

Methods and results: In a 26-week controlled dietary intervention study, 146 subjects followed either NND, a predominantly organic diet high in fruit, vegetables, whole grains, and fish, or ADD, a diet higher in imported and processed foods. Fasting plasma samples are analyzed with untargeted ultra-performance liquid chromatography-quadruple time-of-flight. It is demonstrated that supervised machine learning with feature selection can separate NND and ADD samples with an average test set performance of up to 0.88 area under the curve. The NND plasma metabolome is characterized by diet-related metabolites, such as pipecolic acid betaine (whole grain), trimethylamine oxide, and prolyl hydroxyproline (both fish intake), while theobromine (chocolate) and proline betaine (citrus) were associated with ADD. Amino acid (i.e., indolelactic acid and hydroxy-3-methylbutyrate) and fat metabolism (butyryl carnitine) characterize ADD whereas NND is associated with higher concentrations of polyunsaturated phosphatidylcholines.

Conclusions: The plasma metabolite profiles are predictive of dietary patterns and reflected good compliance while indicating effects of potential health benefit, including changes in fat metabolism and glucose utilization.

Trial registration: ClinicalTrials.gov NCT01195610.

Keywords: LC-MS; biomarker patterns; compliance; plasma; untargeted metabolomics.