Validity of a Food-Frequency Questionnaire in assessing dietary intake in the Southern Cone of Latin America. Preliminary Results

ELORRIAGA NATALIA; IRAZOLA VILMA; BRITZ MÓNICA; MARTINEZ OAKLEY SOLANGE; SUÁREZ EYTEL ROCÍO; GIBBONS LUZ; BAZZANO LYDIA; CORMICK GABRIELA; DEFAGÓ DANIELA; HE JIANG; RUBINSTEIN ADOLFO

Encuentro; NHLBI Centers of Excellence Steering Committee Meeting; 2011

Objective: To adapt and validate the Dietary History Questionnaire (DHQ I) developed by the National Cancer Institute, a semi-quantitative self-administered food frequency questionnaire(FFQ), in adults residing in Argentina, Chile and Uruguay. Population: 150 participants attending primary care centers in the three participating countries. Methods: Observational cross-sectional study. The cross-cultural adaptation process included revision by experts of the three countries and cognitive evaluation of the questionnaire through in-depth interviews. To assess test-retest reliability, the FFQ was administered at baseline (FFQ1) and after 3 months (FFQ2). To evaluate construct validity, trained interviewers performed three 24-hour recalls (DR) by multiple passes to each participant, at baseline, after 3 and 6 months, using photographic models to estimate food quantities.The average daily intake of food, beverages and supplements from both instruments was estimated using a specific software developed ad hoc. The intake of carbohydrates, proteins, total lipids, saturated fats, cholesterol, fiber, vitamins and minerals, as well as alcohol consumption are calculated using a compiled food database developed ad hoc. Test-retest reliability and validity were evaluated by dividing dietary intake in tertiles and calculating weighted kappa (wk) values for each category.Preliminary results: Data from 83 participants, 37.3% men, were included (43 from Argentina, and 40 from Uruguay) with a mean age of 48.8 (SD: 14.9) years old. FFQ1 versus FFQ2: Classifying individual intake estimates into tertiles of the distribution for both FFQs, 50 to 68.7% of the intakes were categorized into the same tertile and 3.6 to 7.2% into the opposite tertile for energy, macronutrients and fiber. The range for wk values was 0.61-0.67 for all the selected nutrients but lipids (wk= 0.53).FFQ1 versus DRs: 44.6 to 48.2% of the intakes of the selected nutrients were categorized into the same tertile and 3.6 to 12% into the opposite tertile. The range for wk values was 0.52-0.68 for all the selected nutrients except protein (wk=0.38).Conclusions: According to this preliminary analysis, the FFQ version for the Southern Cone presents acceptable validity and reliability for its use in epidemiological studies as well as monitoring population dietary patterns and interventions aimed to adopt healthy lifestyles to prevent NCD.