Poster Topical Area: Energy and Macronutrient Metabolism
Location: Hall D
Poster Board Number: 523
Objective: A Mediterranean diet pattern has been associated with health benefits when compared to a Western diet pattern. However, quantifying dietary patterns using self-report of diet may be problematic due to systematic and random errors. Biomarker studies of dietary intake, including metabolomics, may provide a means to characterize dietary patterns, but validation studies are needed. We sought to identify broad metabolomic markers to distinguish Mediterranean and Western dietary patterns in a feeding study of nonhuman primates.
Methods: Urine and blood were collected from 37 female nonhuman primates who were randomized to a Mediterranean or Western diet; the diet had been consumed for 2 years at the time of study. Both diets contained complete vitamin and mineral mixes for primates insuring adequate nutrition, and comparable composition of total protein, carbohydrates and fat. Metabolites were measured using ultra-performance liquid chromatography (UPLC) and a high resolution mass spectrometer. Random forest analysis was used to identify the metabolites that most strongly differentiated the two diet groups. Following log transformation, levels of specific metabolites were compared between groups using Welch's t-test.
Results: The random forest classification analysis of both the plasma and urine metabolic profiles was 100% accurate for differentiating the two groups. For both plasma and urine analysis, metabolites that distinguished the Western diet group from the Mediterranean diet group included those associated with lipid intake and metabolism, amino acid metabolism, diet-derived phenolic metabolites, and micronutrients.
Conclusion: This long-term well-controlled feeding study demonstrated that metabolomics is a promising method of identifying dietary pattern signatures, and warrants further evaluation.
Wake Forest School of Medicine
Winston-salem, North Carolina