Padrões de consumo de alimentos de mulheres: análise dos registros alimentares da Pesquisa de Orçamentos Familiares 2008-2009 e revisão sistemática da relação com câncer de mama

A systematic review of literature about the relationship between dietary patterns identified by factor analysis and breast cancer included studies carried out in different populations from all continents of the world. Were found a dietary pattern compound of vegetables, fruits, fish, shellfish, soy and soy derivatives, olive oil, chicken, and another dietary pattern that described the typical foods habits of the studied region. These dietary patterns seemed reduce the risk of breast cancer. A third dietary pattern characterized by consumption of alcoholic beverages seemed increasing the risk. To investigate the existence of similar patterns in Brazilian women were observed data of individual food intake module of the Brazilian household budget survey (2008-2009) in a subsample of 8,325 non-lactating, non-pregnant and aged over 35 years women. By Exploratory Factor Analysis (EFA) were identified four dietary patterns called “Traditional Brazilian pattern”, Snacks pattern”, “Fats and alcohol pattern” and “Lacto-vegetarian pattern”. The subsample of 8,325 women was grouped in six clusters according to the four dietary pattern observed. Group 1 (n=974 women) was characterized by higher consumption of pattern “Lacto-vegetarian pattern”, Group 2 (n=2,904 women) presented the second highest standardized mean score for the consumption of “Snacks pattern”. Group 3 (n=1,629 women) presented the lowest average consumption of “Traditional Brazilian pattern”. In Group 4 (n=838 women) was observed the high standardized score for the consumption of “Alcohol and fats pattern”. Group 5 (n=1,209 women) showed the highest score for “Snacks pattern” and Group 6 (n=771 women) showed the highest consumption for “Traditional Brazilian pattern” when compared to others. This population performed similarly with respect to distribution through five regions of the country. Were observed significant associations between the six groups and geographic location (urban/rural) (p-value <0.001), BMI (p-value <0.001) and education (p-value <0.001). An inverse correlation was observed between age and education for these women.