Questionário de frequência alimentar ELSA-Brasil : proposta de redução e validação da versão reduzida

• Autor(a) principal: Mannato, Livia Welter
• Data de Publicação: 2013
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Espírito Santo (riUfes)
• Texto Completo: http://repositorio.ufes.br/handle/10/5682
• topic: Diet; Questionnaires; Validation studies; Epidemiologic studies; Estudos de validação; Estudos epidemiológicos; Saúde Coletiva; Dieta; Questionários; Epidemiologia - Estudos longitudinais; 614

Introduction. The FFQ has been the instrument of choice for obtaining dietary information in epidemiological studies, especially those related to NCDs. Many QFAs have been prepared for the national epidemiological surveys, seeking review of the habitual diet of the population being studied, but most of these studies has an extensive list of foods that tend to have higher rate of non-responde, raising the time and cost studies. Objective. Develop reduced version of the FFQ ELSA-Brasil and study the relative validity of the FFQ reduced version. Methods. From the FFQ ELSA-Brasil with 114 food items was held to reduce the list by multiple linear regression, frequency of consumption and nutrient composition of foods. We conducted sensitivity analysis and kappa in which we evaluated the concordance between the methods (FFQ original and FFQ reduced) by classifying individuals into tertiles according to the distribuition of energy and nutrients. Were presented the sums of percentages of concordance exact and adjacent . To assess the validity of the values of energy and nutrients were deattenuated food records and variables of both methods (FFQ and food records) were log transformed, we applied Pearson correlation, intraclass correlation, and sensitivity and kappa to assess agreement. To evaluate differences and biases of energy and nutrients obtained between methods (FFQ - food records) were constructed graphs with absolute differences between the values in the y-axis and the average estimated intake by three food records in the abscissas. Results. Was obtained by reduction of approximately 33% of the original food list. Thus, the new list contains 76 food items. With respect to the validation intraclass correlation coefficients ranged from 0.17 (selenium) to 0.66 (calcium). Found the sum of the values of most adjacent exact distribution above 60% agreement between the methods and food records. Conclusion. Given the results of this study, it is concluded that its objectives were achieved. The FFQ ELSA-Brasil was reduced from 114 to 76 food items, reduction of approximately 33%, and it was still possible to maintain the ability to measure relatively good energy and selected nutrients, obtaining excellent agreement and correlation with the original FFQ.