In-person vs telephone-administered multiple-pass 24-hour recalls in women: Validation with doubly labeled water

Objective To determine the accuracy of energy intakes estimated with the multiple-pass 24-hour recall method in women by conducting in-person and telephone interviews. Doubly labeled water measurements of total energy expenditure were used for validation. Subjects Thirty-five weight-stable women (mean age = 30 years, range = 19 to 46 years) participated. Design Total energy expenditure was measured over a 14-day period using the doubly labeled water method. During this time, 4 multiple-pass 24-hour recalls were obtained from the women (2 in-person, 2 by telephone) who were provided 2-dimensional food models to estimate portion sizes. The Food Intake Analysis System was used to analyze recall data. Statistical analyses Paired t tests were conducted to examine differences between energy intake estimated from the telephone and in-person interviews. Agreement between the energy intake estimates from the telephone recalls and the in-person recalls was assessed using the technique of Bland and Altman. Paired t tests were used to compare energy intake estimated from the telephone and in-person recalls to total energy expenditure. Results No significant difference in mean daily energy intake was found between the telephone (2,253 +/- 688 kcal) and in-person (2,173 +/- 656 kcal) interviews (P = .36). However, the mean energy intake from each interview method was significantly lower than total energy expenditure (2,644 +/- 503 kcal) (P = .006 and .001, respectively). Applications/conclusions Underreporting of energy intake was widespread in the sample. Although the multiple-pass 24-hour recall method did not generate a group measure of energy intake that was accurate or unbiased, the telephone-administered multiple-pass 24-hour recall was just as effective in estimating energy intake as the recall administered in-person. Dietetics professionals should be aware of the pervasive and serious problem of underreporting of self-reported food intakes.