OBJECTIVE: To explore the dietary patterns of adult women and to provide information for validating the instrument used. METHODS: A population-based cross-sectional study was carried out in a representative sampling of 1,026 women aged 20 to 60 years living in Southeastern Brazil, in 2003. A 70-item food frequency questionnaire was used to measure food intake. Dietary patterns were identified using principal component analysis. RESULTS: The reliability of factor analysis was assessed through the Kaiser-Meyer-Olkin statistic, that was equal to 0.823, the Bartlett test of sphericity, which reached statistical significance (chi²(1225)=7406.39; p<0.001), and the determinant of the correlation matrix (6,28-4). Five dietary patterns were identified, each one consisting of 10 items (foods), which were grouped as healthy food pattern cost 1, healthy food pattern cost 2, healthy food pattern cost 3, risk food pattern cost 1 and risk food pattern cost 3. CONCLUSIONS: There were identified five dietary patterns among adult women. There was a clear difference in terms of the cost of these patterns. These findings suggest that cost may be one of the determinants of food choice and food intake in this population.
Dietary patterns; Women; Diet surveys; Factor analysis, statistics; Cross-sectional studies
