Full text
| |
| Author(s): |
Michelle Alessandra de Castro
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Faculdade de Saúde Pública (FSP/CIR) |
| Defense date: | 2011-02-04 |
| Examining board members: |
Regina Mara Fisberg;
Dirce Maria Lobo Marchioni;
Edna Massae Yokoo
|
| Advisor: | Regina Mara Fisberg |
| Abstract | |
Background: The precision and reliability of estimates of food intake are the major challenges in studies about diet and health, since dietary measurement is prone to errors and the human diet exhibits great variation. One of the main components of dietary variability is the within-person variation, which represents the day-to-day variation of intake. From a statistical point of view, the within-person variance is an important source of errors in analysis and interpretation of dietary data because it attenuates the measurements of association, such as coefficients of regression and relative risks, and increases the proportion of individuals on tails of the distribution curve. Objectives: To estimate the variance ratio of energy and nutrient intakes and to calculate the number of days of dietary assessment to the estimation of usual nutrient intakes among Brazilian children. Methods: This is a cross-sectional study performed in 2007 that comprised 3150 children (1 to 6 years) attending daycare and preschools in the municipalities of Manaus, Natal, Recife, Brasília, Cuiabá, Rio de Janeiro, Viçosa, São Paulo and Caxias do Sul. In each municipality, 250 children from public schools and 100 children from private schools were included. Socio-demographic and anthropometric data as well as food intake were collected. Food consumption was measured through weighed food record and estimated food record. Multilevel regression models with complex variance structure were fitted to estimate the within- and between-person variances according to age group (1 to 2 years; 3 to 6 years). The within-person variance was modeled as a linear function of the childs age. The number of days was calculated according to the variance ratio of the nutrients and a hypothetical correlation coefficient (r) between observed and usual intake. Results: Total fiber and total fat showed the greatest variance ratios (>6.00), whereas calcium, pantothenic acid, and phosphorus showed the least ratios (1.33 to 7 2.30). Estimating the usual intake of the nutrients (considering r=0.8) requires between 2 and 37 days for children aged 1 to 2 years, and between 2 and 16 days are necessary for children aged 3 to 6 years. It was also observed an increase in the variance ratios of energy and macronutrients according to age of the child. Conclusion: The elevated variance ratios reinforce the need for collection of a great number of dietary measures in Brazilian children. In addition, the results suggest changes in within-person variance according to age (AU) | |
| FAPESP's process: | 08/07549-0 - Estimative of within and between-person variation of nutrients intake in pre-scholars living in different regions of Brazil |
| Grantee: | Michelle Alessandra de Castro |
| Support Opportunities: | Scholarships in Brazil - Master |