Development of a food frequency questionnaire and a comparison with food records

objective: develop a food frequency questionnaire (ffq) and assess its ability to rank a group of adults according to nutrient intake by a comparison with food records. methods: this study consisted of 2 stages. a) development of ffq. analysis of a food consumption database from a group of 60 adults from communities in the central valley of costa rica led to the identification of 10 food groups which contributed to most of between-person variance in energy and nutrient intake. a quantitative ffq containing 111 foods from the 10 groups and covering 7 days was developed. b) data was collected using the ffq and a 4-day food record from a group of 52 adult inhabitants of an urban community in san josé. energy and nutrient intake as estimated by both methods were compared using pearson's correlation coefficients and student's t-test. results: pearson's correlation coefficients were above 0,4 for 18 of the 22 crude nutrient intakes. the degree of association was lower for energy-adjusted intakes, reflecting underestimation of energy intakes by the ffq. conclusions: when compared with the food record, the ffq was able to rank the individuals according to crude nutrient intake. however, due to underestimation of total energy intake by the ffq, the correlations for the energy-adjusted values were poorer. the ffq requires some modifications in its design and its reproducibility and validity must be demonstrated before being used in epidemiological studies in costa rica.