The objective of this research is to develop and test the validity and reliability of a modified picture-sort food frequency questionnaire between 2 application populations, African-American and Hispanic adolescents. Nutritional patterns developed during adolescence may contribute to obesity and increase the risk for several important chronic diseases later in life, including cancer. With emerging nutrition interventions developed for youth, there is a need for accurate dietary methods to assess the success of these interventions. Traditionally, food frequency questionnaires (FFQs) have been developed and tested in white adults. Since diet is not uniform and varies significantly among different populations, results of studies among white adults, as well as the FFQs themselves, are not readily generalizable to other populations. Considering the limited studies and generally poor to modest results of FFQ validation studies among African-American and Hispanic adolescents, there is a necessity as well as obligation to develop and evaluate methods in minority youth, who have been typically under-represented and under-served in these studies. The picture-sort is a novel FFQ, where foods and drinks are presented as pictures with descriptions on cards and sorted by subjects into piles corresponding to consumption over a given period of time. It is an engaging approach, which requires minimal literacy on the part of subjects compared with a typical pencil and paper food frequency questionnaire. Ethnic and age variations in diet can be taken into account by adding foods and drinks specific to the population as new cards or incorporating them into existing categories. For this study, we will design application sets of cards specific to each of the adolescent populations and test validity and reliability of the approach separately among each sample, African-American subjects will be assessed in year 1 and Hispanic subjects in year 2. Subjects will be recruited through application high schools in Denver. For validity (N=80 for each population), energy and nutrient values from administration of the picture-sort FFQ will be compared with those from the mean of three 24-hour dietary recalls. For test-retest reliability, we will select subjects who have not taken part in the validity study (N=80 for each population) and compare energy and nutrients for application administrations of the picture-sort FFQ conducted application weeks apart.
