Adaptation and validation of an FFQ for 6–10-year-old children

Objective To adapt and test the relative validity of an instrument measuring the usual food intake of 6–10-year-old children. Design An FFQ encompassing the preceding 6 months was adapted and compared with the average of three 24 h dietary recalls. Setting Private and public schools in Porto Alegre, capital city of Rio Grande do Sul, the southernmost state in Brazil, with 1·5 million inhabitants. Subjects Children aged 6–10 years attending grades 1–4 in private and public schools in Porto Alegre. Results Ninety-one children were studied. The FFQ overestimated all nutrients. Correlations with the values obtained by 24 h dietary recalls were mostly above 0·50. The deattenuated correlations increased for all nutrients. The κ coefficients for the adjusted nutrients varied from 0·12 (weak) to 0·34 (reasonable). Graphically, the FFQ was shown to underestimate some of the parameters and to overestimate others, with a wide CI for all nutrients. Conclusions The FFQ does not have the required relative validity to classify the intake levels of schoolchildren, and further investigation is required to understand the sources of error.