Poster Topical Area: Nutritional Epidemiology
Location: Hall D
Poster Board Number: 748
Objective: To analyze differences in nutrients from different food composition databases (DB).
Methodology: Within the framework of a larger project on the environmental footprint of food waste, the nutrient differences in five food composition DB have been analyzed. DB from one country in each of the WHO's global regions have been selected: United States of America (Region of the Americas), Tanzania (African Region), Spain (Europan Region), India (South-East Asia Region) and New Zealand (Western Pacific Region). The three most consumed foods according to the FAO have been compared: milk, wheat and rice. Water, lipids, proteins, carbohydrates, vitamin C and sodium were analyzed according to the denomination, availability, value and units. The Coefficient of Variation (CV) was calculated to estimate the variability of nutritional values.
Results: All nutrients of the three foods were available in the databases studied, except for the percentage of water in Tanzania's DB that could only be estimated indirectly. Nutrients were expressed in all cases in the same units, grams for macronutrients, or milligrams for micronutrients, for every 100 grams of edible portion. The Indian DB was the only one to offer standard deviation of nutritional values, revealing their approximate nature. The nutritional composition of the inedible portion was not available in any case. Apart from language differences, there was uniformity in the denominations. The value of all nutrients showed variations due to the origin of the foods and the chemical analysis used to obtain them. The greatest variation was found in the values of vitamin C in milk (CV = 1.02) and sodium (CV = 0.91) and lipids (CV = 0.33) in rice. The food with the least variability in its nutrients was wheat (CV < 0.26 in all nutrient values).
Conclusion: The greatest differences were found in micronutrients. Dietary calculations made from DB should always be understood as approximate. Global unification of measurement methods and data is recommended. Future research should study data on inedible portions, important for food waste research, both at the household level and in other parts of the food chain.
University of Cantabria
Santander, Cantabria, Spain