Mortality rates for colorectal cancer show regional variations in Brazil, and the heterogeneous dietary profile observed in its geographic regions could possibly explain, at least partially, these different patterns. An ecological study was conducted to identify possible associations between dietary consumption profiles and age-standardized mortality rates for colorectal cancer in selected state capitals. The dietary profile in each capital was characterized on the basis of the consumption of groups of foods associated with this neoplasia in the literature and data on food consumption was obtained through the National Study on Familial Expenditure (ENDEF). Multiple linear regression was used to analyze the correlations between mortality rates and dietary variables. These rates showed a positive correlation with the consumption of calories, cereals, meats, eggs/milk, oils/fats, vegetables/fruits. When we adjusted for total caloric consumption, meat and vegetable/fruits were the only variables that maintained a positive correlation. The final multivariate model with those variables was capable of explaining 92% of the variation in mortality rates in the state capitals. These results suggest that differences in total caloric consumption and in meat and vegetables/fruits consumption could partially explain the different patterns of distribution of colorectal cancer mortality in Brazil.
Colorectal cancer; Dietary consumption; Mortality; Brazilian state capitals; Ecological study
