Development and testing of a scale to evaluate diet according to the recommendations of the Dietary Guidelines for the Brazilian Population

OBJECTIVE: To develop and test a scale for healthy eating practices measurement according to the Dietary Guidelines for the Brazilian Population recommendations. DESIGN: Methodological study. The current Brazilian food-based dietary guideline highlights the importance of choosing foods, combining foods to create meals and modes of eating. These recommendations formed the main domains of the scale and served as a basis for the development of ninety-six items, each with a 4-point Likert response option. Content and face validity were tested. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were performed to determine construct validity. Internal consistency was determined using α and ω coefficients, and reproducibility was tested using test-retest. SETTING: Brazil.ParticipantsA ten-member expert panel was used to assess content validity. Adults aged 18-60 years were included in the face validity (n 20), EFA (n 352), CFA and reliability tests (n 900). RESULTS: Of the ninety-six initial items, twenty-four were excluded and fifty-five were reworded following the content and face validations. EFA detected a four-domain structure (Food choices, Modes of eating, Planning and Domestic organization), which explained 41 % of the variance. CFA led to a final twenty-four-item model with acceptable goodness-of-fit indices and good reliability measures (α=0·77; ω=0·83). Intraclass correlation coefficient for the total score (0·82) and analysis of the Bland-Altman plot suggested good reproducibility of the scale. CONCLUSIONS: The scale presents good evidence of validity and reliability. This innovative study created a useful tool for evaluation of the impact of the Dietary Guidelines for the Brazilian Population.