The Diet Assessment Research Unit is a collaborative resource managed by Cancer Prevention and Control Program senior researcher Tom Hurley that is available to researchers both within the University of South Carolina community and at other institutions and was formed here at USC in 2000. We at the Unit provide expertise in the design, collection, processing and analysis of information on dietary habits and nutrient intake. We assess diet for two main purposes: 1) in observational studies to assess the role of diet in health; and 2) to evaluate the effectiveness of intervention strategies in modifying diet. For the former, we rely mainly on custom-designed food frequency questionnaires that can focus on the overall diet or some subset of foods or nutrients. The goal is often to capture information on usual diet. For interventions, where the intent is to accurately measure current diet, we may use a structured assessment instrument such as the food frequency questionnaire, but more commonly, we will assess individual nutrient or food-group intake using the 24-hour recall interview format. In addition, we also use short-list screeners, which target a specific set of foods or nutrients, and food diaries.
The sections below more fully describe each of our diet assessment methods:
The food frequency questionnaire is the mainstay of work in observational epidemiology. We design instruments to capture specific nutrients and modify them to conform to the specific needs and requirements of study populations. We have extensive experience with developing and testing a range of structured questionnaires, including food frequency questionnaires, used in a variety of research and practice settings. In addition to developing and testing these instruments, we have identified ways to improve precision and accuracy in epidemiologic studies by identifying relevant biases and developing methods of control when performing statistical analyses. A single administration of a food frequency questionnaire provides an estimate of usual, habitual intake. Administration takes about 30-45 minutes to complete and can be either self-administered or interviewer-administered. Output datasets can consist of either food or nutrient data, depending on the needs of the researcher.
24-hour diet recall interviews are used mainly in interventional research and can be administered in person or over the telephone. We use the Nutrition Data System for Research, a state-of-the-art interview and food and nutrient database system developed by the Nutrition Coordinating Center at the University of Minnesota. Registered dietitians trained in the multi-pass interviewing technique conduct the interviews, which take approximately 20-30 minutes. A single day is adequate for estimating the diet of a group of people, but a minimum of 3 days is necessary to estimate nutrient intake for an individual for the most easily characterized nutrients. Additional interviews may be required depending on the nutrient or food group of interest. With a sufficient number of recall days, the 24-hour diet recall interview is the assessment method with the lowest total measurement error. There is a great variety in the types of output datasets that can be created. They consist of nutrient or food data, which can be disaggregated by day or meal type. Datasets can be output in any format the client specifies.
In some research settings, a short, list-based instrument is required to assess dietary habits with a minimum of time and respondent input, with the National Cancer Institute's Fat and Fruit and Vegetable Screening Instruments and other short-list food frequency questionnaires filling this need. We have experience working with the National Cancer Institute-developed screeners and other short-list instruments. These instruments can be self-administered or conducted by an interviewer.
Our Research Unit can provide expert consultation on study design, diet assessment strategies and the statistical analysis of dietary data. Areas in which our group has extensive experience include the design and validation of dietary assessment instruments, with an emphasis on identifying and adjusting for self-report biasers.