Validation of Web-based Multimedia Dietary Assessment Tools

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): There is an increasing demand for low cost and accurate methods to measure dietary intake and physical activity in epidemiologic studies. This proposal requests funds to examine the precision of two new dietary assessment tool, the Computer-Assisted Self-Interview Diet History (CASI-DH), which is modeled after the strengths of the diet history, one of the strongest dietary assessment approaches and a fully automated multimedia 24 hour recall: DIETDAY. The CASI-DH is a web-based, self-interview approach that is meal-, audio-, and food picture-based; generates nutrient and food intake estimates; has an extensive component to assess dietary supplement use and provides feedback to help people adopt more healthful dietary habits. It differs in many significant ways from food frequency questionnaires in that it is fully quantitative, includes thousands of foods, is meal-based, and does not truncate foods or frequencies of consumption. The objective of the validation study proposed here is to test how accurately these new tools capture usual diet by comparing self-reported nutrient and food intakes obtained through the CASI-DH and using the DIETDAY to objective validation biomarkers in a sample of 150 African American and 150 Caucasian adults. Specifically, we aim to examine the validity of both instruments for intakes of nutrients (total energy, protein, and potassium), vegetables and fruits and supplement use by using doubly labeled water for total energy; 24-hour urinary nitrogen for protein; urinary potassium for dietary potassium; and blood levels of nutrients (carotenoids and tocopherols) for fruits, vegetables, and nutritional supplement use. A number of secondary aims will also be pursued, including validation of two physical activity questionnaires and two different accelerometry approaches to DLW. To study the reproducibility of the CASI-DH, DIETDAY and DLW, repeat measures will be conducted on a subset of 100 persons 6 months apart. Due to its low cost, the language independence of the system, and the reliance on images for accurate communication, the CASI-DH and DIETDAY present unique opportunities for epidemiologic research in various populations for assessment of complex dietary exposures. Successful validation of either the CASI-DH or the DIETDAY will represent important methodological advances in nutritional epidemiology. The study will also contribute to our confidence in these dietary and activity assessment instruments in multiracial populations.