The relationship between diet and disease is a critical one in public health. Many studies rely on elementary school children in upper-grade levels to self-report dietary intake. Results from our 2009 validation study showed that 4th-grade children's dietary recall accuracy was affected by retention interval (elapsed time between to-be-reported meals and the interview). Accuracy for the school-meal parts of 24-hour dietary recalls (24hDRs) was better for recalls about the prior 24 hours (the 24 hours immediately preceding the interview) than for recalls about the previous day (midnight to midnight of the day before the interview). However, even when the retention interval was shortest and accuracy was best, recalls had substantial errors: 40% of items observed eaten were omitted (not reported), and 25% of items reported eaten were intruded (not observed). More research is needed to improve children's dietary recall accuracy. Two prominent 24hDR protocols are the Nutrition Data System for Research (NDSR) and the Automated Multiple-Pass Method (AMPM);a new protocol is the Automated Self-Administered 24hDR (ASA24). Software for each of these protocols targets the previous day. Recall accuracy may also be influenced by prompts used to ask subjects for their initial recall of the target period's events. These prompts can be forward (most distant to most recent), reverse (most recent to most distant), open (free [no instructions]), or meal name (breakfast, lunch, etc). NDSR software uses forward prompts, AMPM software uses open prompts, and ASA24 software uses meal name prompts. Any 24hDR has a retention interval and prompts;thus, it is necessary to identify the combination of retention interval and prompts that maximizes recall accuracy, but empirical evidence is limited to a small pilot study. This 4-year project has 3 aims and data collection with 4th-grade children.
Aim 1 is to compare dietary recall accuracy for 8 conditions defined by crossing 2 retention intervals (prior-24-hour recall obtained in the afternoon;previous-day recall obtained in the morning) with 4 prompts (forward;reverse;open;meal). Each of 480 children will be observed eating school breakfast and school lunch, and then interviewed to obtain a 24hDR using 1 of the 8 conditions, with 60 children (30 per sex) per condition. Each child's reported data for the school-meal parts of the 24hDR will be compared to observed data for the 2 school meals, and the child's accuracy will be assessed for food items, amounts, energy, and macronutrients.
Aim 2 is to assess relationships of social desirability, body mass index, socioeconomic status, and achievement test scores with dietary recall accuracy of 480 children in Aim 1.
Aim 3 is to conduct 1-month test-retest reliability (in Year 1) for a 14-item social desirability scale via classroom administration to approximately 210 children. This R01 project will compare crucial yet untested aspects of 24hDR protocols and provide empirical evidence for refining software to obtain more accurate 24hDRs from children for epidemiologic studies, interventions, and clinical practice, and thereby improve our understanding of relationships between diet and disease.
The relationship between diet and disease is a critical one in public health, but dietary assessment is challenging, especially among school children. Misreporting in dietary surveys is a widely recognized problem that threatens our ability to pinpoint true relationships between diet and disease. This project will provide empirical evidence for refining software to obtain more accurate 24-hour dietary recalls from children for national surveys, epidemiologic studies, interventions, and clinical practice, and thereby improve our understanding of relationships between diet and disease.
