Relative Validation of a Four Weeks Retrospective Food Frequency Questionnaire versus 7-Day Paper-Based Food Records in Estimating the Intake of Energy and Nutrients in Adults

Abstract

Purpose: This study assessed potential differences in estimating short-term dietary intake of energy and nutrients and food consumption, between 4-week food frequency questionnaires (FFQs) and 7-day food records in Romanian adults. Patients and Methods: A total of 116 participants (age range 18–74 years, 31% males and 28.4% of participants being overweight and obese) were recruited. Estimates for energy and macro- and micronutrient intakes, and food group intakes were compared between the two methods using Wilcoxon-sign-rank test, correlation coefficients, Cohen’s Kappa, Bland– Altman plots with 95% limits of agreement, and quartile classifications. Results: Cohen’s Kappa values for energy and macronutrient intakes indicated moderate agreement, ranging from 0.402 (protein) to 0.470 (fat), fair agreement for most micronutrients (0.2–0.4) and poor agreement for most food groups (<0.2). When data were cross-classified into quartiles for energy and macronutrients, about 58% of participants were cross-classified in the same quartile using both methods, while 33% of participants were cross-classified in adjacent quartiles of one method versus the other. Micronutrients (such Na, Mg, Ca, K, Fe, vitamins) had the highest degree of misclassification, on average 40% being cross-classified in the same quartile and another 40% in adjacent quartiles. Bland– Altman plots suggested that both methods were comparable for energy and all macronutrients. When the consumption of food groups was compared, correlation coefficients between methods ranged from 0.09 (legumes) to 0.26 (whole grain), indicating poor correlation. Conclusion: These results showed that the relative match of a standard FFQ, as compared to the 7-day food records, was moderate in estimating macronutrient and energy, fair for most micronutrient intakes and poor for others and as for food groups.