Accuracy of a 24-Hour Recall of a Measured Snack (P13-005-19).

Abstract

Objectives\nTo assess food recall accuracy in the general population and determine recall differences in persons with and without food addiction (FA) as identified by the Yale Food Addiction Scale. We hypothesized that food recall is inaccurate for individuals regardless of FA status. Furthermore, subjects with FA will have a more accurate food recall compared to those without FA.\n\n\nMethods\nTwenty-two individuals pair-matched based on gender, race, age, and BMI, participated in this cross-sectional study (11 FA and 11 control). In the outpatient clinic setting, subjects were offered ad libitum a highly palatable snack (Lays™ potato chips and plain M&M s™) approximately 3 hours after drinking a meal replacement shake. The snacks were covertly weighed before and after consumption to note the true intake. A link to an ASA-24 hour dietary recall was emailed to the participant the following day and the information collected was compared with the true measured intake of the snack.\n\n\nResults\nAll participants under-reported an average of 3 g of M&M sTM and 6 g of Lay sTM potato chips, corresponding to 15 kcal and 34 kcal (total of 49 kcal). Overall, subjects under-reported M&M sTM intake by 14.3% and Lay sTM intake by 25.5%. Persons with FA under-reported M&M sTM intake by 6.8% and Lay sTM intake by 24.8%; the controls under-reported M&M sTM intake by 21.9% and Lay sTM intake by 26.3%. Differences in reporting accuracy were not significantly different in the paired sample.\n\n\nConclusions\nReporting of 49 fewer calories of a 169 calorie snack can be considered substantial underreporting. Individuals appear unable to recall food intake accurately even when recalled within 24 h of consumption. FA does not seem to influence overall food recall accuracy differently. Moreover, the % variations in underreporting between the two varieties of snacks, and between FA or non FA groups suggests correcting underreporting by a single correction factor is not a viable option. A larger sample size would assist in the accuracy of conclusions.\n\n\nFunding Sources\nTexas Tech University.