Models of Underreporting of Total Energy Intake on the Diet History Questionnaire and 24-hour Recalls

• Activity Number: 39
• Type: Contributed
• Date/Time: Sunday, August 11, 2002 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistics in Epidemiology*
• Author(s): Janet Tooze*+ and Victor Kipnis
• Affiliation(s): National Cancer Institute and National Cancer Institute
• Keywords: nutrition ; underreporting ; diet ; food frequency questionnaire

Abstract

Self-reported measures of diet contain a large amount of measurement error, which usually leads to underreporting of usual intake at the population level. Using doubly labeled water (DLW) to measure total energy expenditure and the Diet History Questionnaire (DHQ) and two 24-hour recalls (24HR) to measure self-reported dietary intakes, models to predict underreporting were developed using data from 484 men and women ages 40-69 yrs from the Observing Protein and Energy Nutrition (OPEN) Study. Numerous factors collected in OPEN were considered in multiple logistic regression models to predict the probability of being an underreporter. In DHQ models, weight loss, fear of negative evaluation, and percent of calories from fat were the best predictors of underreporting in women; body mass index, comparison of activity level to others, and eating frequency were the best predictors in men. In 24HR models, psychosocial factors were also related to underreporting, as well as dieting, percent of calories from fat, and variability in diet. The models explained a low percentage of the variability associated with underreporting, especially on the DHQ.