The goals and objectives of this study are: l) To test the hypothesis that correlations with nutrient intake estimated by multiple 24-hour recalls will be greater for an innovative, cognitively-developed food frequency questionnaire (FFQ) than for two commonly used FFQs (the NCI-Block Health Habits and History Questionnaire (HHHQ) and the Harvard-Willett Questionnaire (WQ)). 2) To test the hypothesis (in a 20% subsample of respondents) that the correlations between plasma carotenoid levels and the new FFQ are higher than those obtained using either the HHHQ or the WQ. In order to accomplish these goals, a study is being conducted in a defined free-living population in which the contractor will collect or provide the following data: 1) FFQs and 24-hour dietary recalls administered over the course of one year. 2) nutrient intakes from all 24-hour dietary recalls and FFQs. 3) blood specimens which will be analyzed for plasma carotenoids and lipids in a 20% subsample. Following the contract, the NCI investigators will perform all statistical analyses of correlations between the FFQs, 24-hour dietary recalls, and blood measures.
