Lecciones aprendidas en un estudio de reproducibilidad y validez de cuestionarios para estimar la ingesta dietética en Colombia. 2010-2017

Introduction: Dietary exposure is complex to measure and is associated with the development of chronic diseases; therefore, it is necessary to develop and validate simplified instruments for measurement. Objective: To establish the classification and validity reproducibility of the 24-hour dietary recall method (24hDR) in food frequency questionnaires (FFQ) developed in Colombia. Materials and methods: Diagnostic technology study developed during 2010-2017. Two 24hDR and two FFQs were applied to 523 subjects: 109 in Bogota D.C., 107 in Medellin, 106 in Cali, 102 in Barranquilla and 99 in Bucaramanga and its metropolitan area. Reproducibility and validity studies were performed using crude Pearson’s r coefficients adjusted by residual energy method. Results: The reproducibility of 24hDR was poor-moderate (minimum -0.00, maximum 0.72), as was the reproducibility of the questionnaires (minimum -0.01, maximum 0.72). The validity of the questionnaires was poor (minimum -0.01, maximum -0.59). Conclusion: FFQs are not useful for epidemiological research, which is explained by the high variability caused by few 24hR recalls in dietary estimation when designing questionnaires, and failures in data collection.