RESUMO
AIM: The present study aimed to validate food records on the application MyFitnessPal (MFP), comparing them with paper-based food records (P-FR). METHODS: Thirty university students, including males and females, volunteered and recorded dietary intakes on P-FR and MFP food records (MFP-FR). The values of energy, macronutrients and fibre from MFP-FR were compared with data from P-FR, calculated using Brazilian food composition tables. Adjustments for in-person variability and energy intake were performed, and comparisons were made between each data set, using the Wilcoxon signed-rank test, Spearman's correlation and Bland-Altman agreement plots. RESULTS: Positive moderate correlations between P-FR and MFP-FR for all variables, and non-significant associations for energy and fibre were found. The Bland-Altman plots showed tendency to underestimation and relatively narrow limits of agreement. Carbohydrate and lipids show trends of increasing the degree of overestimation with increased intake, even after data normalisation. CONCLUSIONS: MFP tends to underestimate ingestion of nutrients probably due to inadequacies in the MFP database. However, MFP showed good relative validity, especially for energy and fibre. Its use, as well as other similar applications, should be encouraged, due to ease of assessing dietary information, although careful usage is recommended because of database gaps.