The National Cancer Institute's automated self‐administered 24‐hour dietary recall (ASA24)

Abstract

Extensive evidence indicates 24‐hour recalls (24HR) provide high‐quality dietary data; multiple administrations can be used to model usual intakes. Traditional 24HR, however, are expensive and impractical for large‐scale research as they rely on trained interviewers. To address this, we developed ASA24, a web‐based tool that enables multiple, automated, self‐administered 24HR. It consists of a Respondent website to collect recall data and a Researcher website to manage study logistics and obtain analyses. The Respondent website is modeled on USDA's interviewer‐administered Automated Multiple Pass Method. Unique to the Respondent site are multi‐level food probes to obtain details on food types and amounts. A dynamic user interface, including a talking, animated guide, facilitates finding foods and beverages, jogs respondents’ memories, and assists in estimating portion sizes from multiple images. Resulting data files include food code, nutrient, and food group variables for each day and food reported. Optional modules collect data on location of meals, who one ate with, TV/computer use during meals, and supplement intake. ASA24 will be compared to interviewer‐administered 24HR in a large sample of adults and validated in a small feeding study. Converting interviewer‐administered 24HR to a self‐administered tool makes the collection of high‐quality dietary data feasible and economical. Funding: NIH.