Shared Left Border
    Main Entities
    Estimating Values
    Recipe Calculation
     -  Calculations vs. Analyses
     -  Moisture Adjustment
     -  Weights and Measures
    Data Presentation
    Data Interchange
    Site Overview
    DFI Home
    SciName Finder™

Calculated versus Analytical Nutrient Values

 Updated 2020-09-29

 How accurate are the calculated values compared to analytical values?

In food composition tables and databases many values are calculated. This holds for nutrient values calculated from other nutrients, as well as for the nutrient values for whole foods calculated from other foods or ingredients. For example, the nutrient protein is commonly calculated from so-called Kjeldal nitrogen, an analytical method to determine the total content of nitrogen in a (food) sample. The protein content is the calculated from the total nitrogen by multiplying the nitrogen content with a factor. In many (/most) cases this will lead to an overestimation of the protein content, either due to a considerable amount of non-protein nitrogen in the food or simply because the used nitrogen-to-protein factor is too high.

Similarly, in recipe calculations the nutrient value of whole foods is calculated from the foods ingredient(s) undertaking precautions of adjusting moisture and fat content (yield) as well as destruction of nutrients (nutrient retention) under the conditions of the preparation of the food. For example, simple physical disintegration of a food like cutting cabbage or salad will lead to destruction (oxidation) of some of the vitamin C content, storage of fruit will similarly lead to a reduction of vitamin C. Some nutrients are heat sensitive and will destruct during heat treatment. This means that any treatment the foods or its ingredients undergo during preparation will definitely change the nutrient values of the final food.

Therefore, recipe calculations are not only simple mass balances, but actually sometimes rather sophisticated calculations taking into account the yield after preparation as well as correcting for the supposed loss (or uptake) of vitamins and minerals during preparation. However, it should be catirized as "the best guess available".

The alternative to recipe calculations is analysing all the needed foods for the nutrients needed - a very expensive and time consuming affair - and it is utopia to think that it is feasible.

Recipe calculation is a strongly needed too in food composition and food consumption work. It is therefore necessary to know how good estimates the recipe calculations deliver. It is evident that the calculations cannot be better that the analytical values. It is also evident that it is necessary for the data for ingredients involved to be reliable (good quality) and match the ingredients originally used for the food. Similarly, the nutrient retention factors used must reflect the food and the conditions the food is prepared under.

So, how good are the recipe calculations to mirror the real foods nutritional properties? - the scientific literature is not overwhelming with studies of comparisons of analytical versus calculated nutrient values, but there a few interesting studies - and the results are diverse and ambiguous:

For an updated list of papers on calculated versus analyzed nutrient values, try this Google Scholar search or this Google Scholar Search.

Swedish Food Composition Database

A new version of the Swedish Food Composition Database released by the National Food Administration, see Livsmedeldatabasen.
FoodData Central
October 2020 release.

A new version of the USDA ARS FoodData Central was released October 2020.
For more information, see FoodData Central website.
The French Ciqual 2020 table is online

With 3,185 foods and 67 constituents, the 2020 version of the Ciqual table is available. See the ANSES-CIQUAL website.
First Anniversary release of USDA ARS FoodData Central

For more information, see USDA National Agricultural Library's Food and Nutrition Updates - May 2020.