Aim: To assess the validity and the short-term reproducibility of a semi-quantitative, self-administrated food frequency questionnaire and a structured 72-hour recall in a Mediterranean Spanish population. Methods: 44 free-living volunteers participated in the study. Macronutrient, vitamin and mineral intake, recorded on the food frequency questionnaire and the structured 72-h recall were compared with intakes derived from a three-day food record (reference method). Validity of the dietary assessment methods was further assessed by comparing urinary nitrogen, plasma vitamin C, plasma beta -carotene and whole blood glutathione peroxidase activity levels with the corresponding nutrient intakes from the questionnaires. The food frequency questionnaire and the 72-h recall were administered twice to assess the short terra reproducibility. Results: Pearson's correlation coefficients between urinary nitrogen, plasma vitamin C, plasma beta -carotene and whole blood glutathione peroxidase activity levels and the reported nitrogen, vitamin C, beta -carotene and selenium intakes were 0.26, 0.53, 0.17, 0.26 for the food frequency questionnaire; 0.41, 0.09, 0.34, 0.42 for the structured 72-h recall and 0.50, 0.54, 0.44, 0.38 for the three-day food record, respectively. The short term reproducibility of analysed nutrient intake showed average intra-class correlation coefficients of 0.91 and 0.69 for the food frequency questionnaire and the structured 72-h recall irrespectively. Conclusion.-The food frequency questionnaire and the structured 72-h recall provide valid estimates of nutrient intake and could be used for dietary assessments. (C) 2001 Harcourt Publishers Ltd.
