Nutritional epidemiology is concerned with quantifying dietary exposures and the association of these exposures with risks for disease. Diet represents one of the most universal biological exposures;however accurate assessment of food and beverage intakes is problematic. This proposal addresses the challenges facing self-report methods, such as dietary assessment, that are prone to measurement error and other biases. These problems are further compounded by high respondent burden and high researcher burden. The advances being developed through this proposal are directly addressing these challenges with specialized technology and rigorous testing to insure improved reliability and validity. The focus for this proposal will be a novel food record method using a mobile device, with an embedded camera that will provide an accurate account of daily food and nutrient intake among adolescents. Our approach includes the use of digital image analysis for identification and quantification of food consumption based on images taken of the food items. Data from images obtained before and after food is consumed can be used to link the identified foods and amounts eaten with a food composition database. The volume estimates from the segmentation and imaging analysis are reported in cubic centimeters. Current databases do not contain complete density information to translate the volume estimates to grams of food. Therefore, various algorithms for porous food are being developed to predict density. To aid with interaction design, three convenient samples of adolescents ages 11-18 y will be recruited. The first group (n=80) will participate in one or two controlled meal sessions. Prior to the first meal, participants will be instructed to capture an image with the mobile phone food record (mpFR) that includes all foods and beverages, and the fiducial marker (an item of known size), before, during, and after eating. The second group (n=20) will consume meals over a 24-hour period under controlled conditions. The users of the mpFR under these controlled conditions will provide feedback for improving the accuracy and ease of use of the mpFR. A convenient sample of 100 free-living adolescents between 11 and 18 y will participate in the validation phase where total energy expenditure will be measured over 7 days with doubly labeled water and compared to total energy intake over the same 7 days as estimated from the mpFR. This data collection phase will last just over one year and half of the sample will participate for two 7 day periods (1 in the summer and 1 during the school year). The outcomes of this project will ultimately provide scientists with innovative tools that can be used in school-, population-, and clinical-based studies to provide accurate dietary intake data. The implications will likely translate to enhanced quality in epidemiologic and clinical nutrition research.
The outcomes of this project will ultimately provide scientists with innovative tools that can be used in school-, population-, and clinical-based studies to provide accurate dietary intake data about adolescents. The implications will likely translate to enhanced quality in epidemiologic and clinical nutrition research.
