Dietary patterns for meals of Brazilian adults.

The use of dietary patterns (DP) in nutritional research is well established; however, only a few studies of DP according to specific meals have been conducted. The purpose of this study was to identify the DP regarding breakfast, lunch and dinner meals of the population (aged 20 years and older of both sexes) that participated in the Health Care Survey of Sao Paulo. Food intake was estimated by using the Multiple Source Method - considering two 24- h dietary recalls. On the basis of the food groups for each meal, a factor analysis, with a principal component estimation, was applied (varimax rotation) in order to derive the DP. Prevalences of meal skipping were 5·6 % for breakfast, 3·6 % for lunch and 12·8 % for dinner. The findings revealed three breakfast DP: healthy, traditional and snack; five lunch DP: traditional, salad, sweetened juice, Western and meats; and four dinner DP: coffee with milk and bread, transitional, traditional, and soup and fruits. The results of this study indicate that the DP identified in accordance with the meal nicely discriminates food intake, emphasising peculiarities that are not found in global analyses and might support dietary advice.