Fig. 1 Networks of consumption of foods in the Potsdam EPIC Study (Food frequency questionnaire, ... more
Printversion of: http://www.www.dife.de/forschung/abteilungen/projektdetail.php?id=258&abt=EPI
Status: 24.06.2019 21:43:54
Investigations of the relation of single foods or food groups and risk of chronic diseases are ongoing activities of the department. We use various data sources such as data of the entire European Prospective Investigation into Cancer and Nutrition (EPIC) Study, data of EPIC- Germany (collaboration with the EPIC group in Heidelberg), and data of the local Potsdam EPIC Study . The departmental interests in this research area focus on understanding the interrelation and pattern formation of food intake considering meals, daily intakes, and eating habits. The detailed EPIC-Soft data from 24-h recalls allow research into circadian food patterns. We could describe the overall circadian pattern of energy, nutrient, and food intake in the Potsdam EPIC cohort and – in collaboration with other cohorts – in Germany, and also the type of foods being eaten at the various meals in the EPIC centers. Meal pattern analyses were also linked with metabolic biomarkers, giving in-sight into the metabolic consequences of certain circadian food habits. The department also provided food patterns on the basis of principal components for the whole EPIC Study, now to be used for the different endpoints of chronic diseases. Network analysis was recently used to identify clustering foods, which should be carefully evaluated regarding risk due to potential confounding (Fig. 1). It was determined whether such food clusters could be retrieved by applying the new variable selection method Random Survival Forest.
In the framework of the German National Cohort a new concept of measuring food intake was developed based on the web application of an easy-to-apply 24-hour food list. The scientific basis of the concept was developed by the department and the instrument is now available as a web-based instrument for broader in-house use and in the dietary community in Germany. Currently, simulation work is performed to improve and substantiate the concept, which also requires complex statistical modeling and food frequency questionnaire data as covariate information.
© 2019 DIfE - German Institute of Human Nutrition. All rights reserved. // as of 29.03.2018