Food Matrix Triangle

Much of the analytical data on the nutrient content of foods is generated using official methods of analysis (e.g. AO AC International). An evaluation of AO AC Methods of Analysis for Nutritional Labelling is available (Sullivan and Carpenter 1993). While these methods have often been studied for a variety of food matrices, applicability over the entire range of food matrices has not been formally studied in most cases. In addition, RMs are not available over the entire range of food matrices (Wolf 

In order to define this variety of food matrices, chemical composition differences that primarily influence chemical analytical measurements have to be considered. Major food components determining basic chemical make-up are the proximate composition of fat, protein, carbohydrate, ash, and moisture. Variations in ash content in general have a minor influence on analytical methods for other constituents and impact of moisture content can be controlled. Thus the major components influencing analytical performance are the relative levels of fat, protein, and carbohydrate. 

Drawing on this representation, an approach has been described to systematically describe selection of food products to evaluate the applicability of collaboratively studied methods over a range of food matrices (Wolf and Andrews 1995). A food matrix is described by its location in one of the nine sectors in the triangle. Foods falling within the same sector are chemically similar and thus should behave in a similar analytical manner. This same scheme can be used to select food matrices representing each sector for development of a series of RMs representing all foods. 

In order to begin this selection process, the entire range of food matrices can be examined. The proximate components of foods are plotted to determine into which of the nine sectors the foods fall. To develop an xjy plot of these sectors, values for x and y are assigned using the following equations based upon the% fat (a), % protein b) and% carbohydrate (c) relative to the sum of these three components, normalizing these components to a dry weight, ash-free basis. Thus the sum of a -i- i) -1- c is 100 %. 

The resulting tables (i.e. Table 6.5) list commonly consumed foods that are near the mean x and y for each sector. Commonly consumed foods are a real advantage when preparing RMs, because they tend to be readily available and ineiqiensive. 

---