Adaptation of a Food Frequency Questionnaire to Assess Diets of Puerto Rican and Non-Hispanic Adults

• Published in: American journal of epidemiology 1998-09, Vol.148 (5), p.507-518
• Main Authors: Tucker, Katherine L., Bianchi, Lisa A., Maras, Janice, Bermudez, Odilia I.
• Format: Article
• Language: English
• Subjects: Aged; Biological and medical sciences; diet; Epidemiologic Methods; ethnic groups; European Continental Ancestry Group; Female; Food; Hispanic Americans; Human physiology applied to population studies and life conditions. Human ecophysiology; Humans; Male; Medical sciences; Middle Aged; Nutrition Surveys; Nutritional survey. Food supply and nutritional requirement; Puerto Rico; questionnaires; Surveys and Questionnaires
• ISSN: 0002-9262; 1476-6256
• DOI: 10.1093/oxfordjournals.aje.a009676

To study issues of diet and health among Hispanic adults living in the northeastern United States, the authors adapted a version of the National Cancer Institute (NCI)/Block food frequency questionnaire. Foods that contributed to nutrient intake of Puerto Rican adults in the Hispanic Health and Nutrition Examination Survey (HHANES) were ranked to identify items to be added to the food list. Portion sizes were compared across HHANES and the Second National Health and Nutrition Examination Survey (NHANES II) to assess the adequacy of the assumed values. Within line items, frequencies of consumption of individual foods were ranked and these data were used to adjust the weighting factors within the database. To test the revised form, 24-hour recalls were collected from 90 elderly Hispanics and 35 elderly non-Hispanic whites. These data were coded into the original and revised food frequency forms and nutrient intake results were compared with recall results by paired r-test, and by Pearson and intraclass correlations. Added foods include plantains, avocado, mango, cassava, empanadas, and custard. Portion sizes differed significantly between HHANES and NHANES II, and were left open-ended. Estimated mean nutrient intakes and correlations with recall data were lower with the original versus the revised form. The authors conclude that the use in minority populations of food frequency questionnaires developed for the general population is likely to result in biased estimates of intake unless modifications are made in the questionnaires. Am J Epidemiol 1998; 148:507–18.