Elimination of the Variance Between Individuals Is Necessary to Evaluate the Impact of Garlic on the Metabolic Profile of Human Urine

Abstract

Abstract Objectives To determine the impact of garlic on the metabolic profile of urine. Methods On the first day 17 fasting subjects were fed a breakfast of bread and butter. Urines were collected before and 3 hours after the meal. On a second day, the same 17 fasting subjects were fed a meal of bread, butter, and garlic. Urines were again collected before and 3 hours after the meal. Samples were analyzed by metabolomics using liquid chromatography-mass spectrometry (LC-MS) and data were subjected to analysis of variance-principal component analysis (ANOVA-PCA). Results 637 compounds were found in the urines and 277 were identified. PCA of urine profiles were dominated by variation between individual. Removal of individual variance by ANOVA allowed differentiation of fasting urines from bread and butter urines from bread, butter, and garlic urines. PCA loadings identified compounds that led to discrimination between treatments. Influence of the loading identified compounds were verified by examination of the LC-MS data for individual compounds. Three unique sulfur containing compounds were identified. Loadings showed, however, that a change in the metabolite profiles (ratios of compounds) and not the unique compounds) were most informative. Conclusions Removal of variance between individuals is essential to properly analyze the data. Changes in the patterns of compounds routinely observed in urine were the major result of the garlic meal. ANOVA-PCA is an excellent tool for isolating experimental factors. Funding Sources Agricultural Research Service, US Department of Agriculture and Office of Dietary Supplements, National Institutes of Health.