Nutritional Redundancy in Human Diet

Abstract

Studying human dietary intake may help us identify effective measures to treat or prevent many chronic diseases whose natural histories are influenced by nutritional factors. Through examining four comprehensive cohorts with dietary intake data collected on different time scales, we find that the food intake profile varies tremendously across individuals and over time, while the nutritional intake profile is highly stable across individuals and over time. We refer to this phenomenon as ‘nutritional redundancy’. We find that this phenomenon cannot be simply explained by the fact that different foods contain similar nutrients. Instead, it is largely due to more sophisticated features (e.g., the highly nested structure) of the food-nutrient network --- a bipartite graph that connects foods to their nutrient constituents. The food-nutrient network also enables us to quantify the level of nutritional redundancy for each diet record of any individual, i.e., the personal nutritional redundancy. Interestingly, this personal nutritional redundancy measure does not strongly correlate with any classical healthy diet scores, but its performance of predicting healthy aging in a cohort of older women is as strong as those healthy diet scores. The concept of nutritional redundancy developed here offers us a new perspective on studying human diet and presumably could be used for more phenotype predictions.