Daily intake information is important for an understanding of the metabolic fluctuation of humans exposed to environmental stimuli. However, little investigation has been performed on the variations in dietary intake as an input and the relationship with human fecal, urinary, and salivary metabolic fluctuations as output information triggered by daily dietary intake. In the present study, we describe a data-driven approach for visualizing the daily intake information on a nutritional scale and for evaluating input-output responses under uncontrolled diets in a human study. For the input evaluation of nutritional intake, we collected information about daily dietary intake and converted this information to numeric data of nutritional elements. Furthermore, for the evaluation of output metabolic, mineral, and microbiota responses, we characterized the metabolic, mineral, and microbiota variations of noninvasive human samples of feces, urine, and saliva. The data-driven approach captured significant differences in the fluctuation of intestinal microbiota and some metabolites caused by a high-protein and a high-fat diet in daily life. This approach should contribute to the metabolic assessment of humans affected by environmental and nutritional factors under unlimited and uncontrolled diets.
Keywords: comprehensive significant test; data-driven approach; dietary intake variations; metabonomics; noninvasive human sampling; nutrient digitalization.