[Design and validation of a self-reported dietary polyphenol intake frequency questionnaire.]

Introduction: although there is evidence linking polyphenol consumption with better health outcomes, information on consumption patterns, particularly in Latin American countries, is scarce. Objective: to design and evaluate the validity of an online self-reported food frequency questionnaire to measure dietary total polyphenol intake in the Chilean adult population. Methods: the designed questionnaire (FFQ-P) was submitted to expert opinion for content validity coefficient estimation; then a pilot study was conducted in 47 adults who completed FFQ-P, in addition to a six-day food record as a reference method. Wilcoxon was applied in related samples, Spearman's correlation to determine the strength of association and Bland-Altman diagrams to examine the difference between methods. Results: content validity reflected concordant values for adequacy (0.94 ± 0.04) and relevance (0.93 ± 0.05). The percentage difference between both methods was 1.2 %, with greater variability observed in fruits (9 %) and alcoholic beverages (4 %). The Wilcoxon test showed no significant differences between food groups. Medium reliability was observed for vegetables, fruits, tea, and coffee (0.26 to 0.5) and moderate reliability for alcoholic beverages (0.5 to 0.75). The mean value of the difference was 9.8, being within a short distance of zero. Conclusion: FFQ-P is reliable for estimating polyphenol intake from fruits, vegetables, alcoholic beverages, tea, and coffee; however, the intake of non-alcoholic beverages, nuts and olives should be interpreted with reservation.