Accuracy of 24-Hour Diet Recalls in Older Korean Women

Abstract

Abstract Objectives Despite the wide use of 24-hour diet recalls, its validity has not been investigated in Korean women. The purpose of this study was to assess the accuracy of 24-hour diet recalls in older Korean women. Methods Twenty-two women (mean age: 55.1 years) participating in a 5-day controlled-feeding study were randomly selected and interviewed. Interviews were conducted according to the 5-step multiple-pass method and the protocol of the Korea National Health and Nutrition Examination Survey. Recalled food items were compared to provided foods and matched as: exact, close, or far. The proportion of foods consumed but not reported (exclusions), and number of intrusions (reported but not consumed) were investigated. Accuracy of recalled foods was also analyzed according to Korean food culture (rice, soup, kimchi, side-dish, sauce, snack, and drink). Duplicate meals were weighed, homogenized, and freeze dried. Nutrient content of recalls and recipes of provided food were analyzed by CAN-Pro 5.0 software. Food composites were analyzed for select minerals by inductively coupled plasma optimal emission spectrometer (ICP-OES). Data were analyzed by paired t-test or Wilcoxon Signed-Rank test. Results Among the foods reported, 74% were exact matches. When analyzed according to Korean food culture, kimchi (91%), followed by snacks (81%), side dishes (74%), rice (67%), and drinks (66%), had a higher proportion of exact matches than soup (59%) and sauces (13%). Sauce also had the highest rate of exclusions (21.6%). Only one intrusion happened during the study period (kimchi). The participants reported significantly lower intakes of sodium (mean difference: −1270.0 mg; P < .0001) and fat (mean difference: −11.5 g; P = 0.0003) compared to actual recipes analyzed by CAN-Pro software. Reported intake of other nutrients were similar to actual recipes. However, phosphorus, sodium, potassium, and iron intakes appeared greater when recipes were analyzed by CAN-Pro compared to analysis of duplicate diets by ICP-OES. Conclusions This study is the first to assess the accuracy of 24-hour recalls in Korean women. Older Korean women may accurately recall their dietary intake. However, the Korean nutrient analysis software may not accurately represent mineral intake. Further research in free-living individuals is required. Funding Sources The Ministry of Oceans and Fisheries, Korea.