Comparison between analyzed and calculated food composition data: carotenoids, retinoids, tocopherols, tocotrienols, fat, fatty acids, and sterols

To evaluate the differences in carotenoid, retinoid, tocopherol, tocotrienol, fat, fatty acid, and sterol data obtained by different procedures, the compositions of three pooled diets differing in their fat source, milkfat (MF), rapeseed oil (RO), and sunflower oil (SO), were estimated by two different databases and by analyzing duplicate portions of the diets. The analyzed beta-carotene, all-trans-retinol, and retinol equivalent (RE) values of both oil diets were only 50-60% of the calculated values. By contrast, the analyzed gamma-tocopherol content for the MF and the SO diets was 120 and 180%, respectively, of the calculated contents, but only 90% in the case of the RO diet. The analyzed values (A1) of total fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids were 100-109%, 96-113%, 84-89%, and 88-99%, respectively, of the calculated values when the new Finnish national food composition database (C1) was used. In the case of sterols the analyzed values (A2) ranged from 70 to 98% of the calculated values (C1). The differences found between analyzed and calculated diet composition might be explained by real differences in food composition, different enrichment procedures for margarines, food preparation losses, differences in the analytical methods used, the processing of the duplicate portion sample, and possible inaccuracies when nutrients present in small concentrations are analyzed. The data support the superiority of the newer database (C1) and the methods used to generate A1 values, as well as illustrates the methodological and database biases. In conclusion, values generated using appropriately procured and processed samples assayed with validated methodology should be the basis for assessing the validity of calculated nutrient levels