Validation of 'POIBA-How do we eat?' questionnaire in 9-10 years old schoolchildren.

Background: It is difficult to obtain good food reports with Food Frequency Questionnaires (FFQ) among children. In addition, validated questionnaires are scarce. Objective: The aim of this study was to validate the 'POIBA-How do we eat?' (POIBA-HDWE) FFQ and whether it could be administered to children under 10 years of age. Design: We validated the FFQ POIBA-HDWE as part of the Childhood Obesity Prevention Program (POIBA project) in Barcelona. Forty-two out of 63 primary school students (9-10 years old) answered three questionnaires: FFQ POBA-HDWE; another questionnaire, 'POIBA-How do our children eat?' (POIBA-HDOCE), which was administered to the children's parents; and the 24-h recall computer program 'Young Adolescents' Nutrition Assessment on Computer' (YANA-C), which was used on three different days as a gold standard. We tested for correlations using the Spearman test for non-parametric variables. Results: We found low compliance with food recommendations (<50%). The POIBA-HDWE and POIBA-HDOCE questionnaires showed a moderate correlation for soft drinks (r = 0.49; p < 0.01), nuts (r = 0.59; p < 0.01), dairy products (r = 0.41; p < 0.01) and juices (r = 0.49; p < 0.01). There were moderate correlations between POIBA-HDWE and YANA-C for fried potatoes (r = 0.42; p = 0.01), dairy products (r = 0.53; p < 0.01), juices (r = 0.41; p < 0.01), and grains(r = 0.50; p < 0.01). Food frequency questions showed a homogeneity of 0.69, and a sensitivity of over 60% for all food items except chips (37.5%) and sweets (51.7%). Conclusions: The POIBA-HDWE FFQ showed moderate correlations with the gold standard, high sensitivity for most food types and acceptable internal consistency. It is an easy and affordable tool for recording food frequency in children under 10 years old.