Variations in major nutrients and nutrient data in Swedish foods: effects of sampling, preparation and determination
1997
Torlem, I. (SLU, Uppsala (Sweden). Inst. foer Livsmedelsvetenskap)
This thesis summarises and discusses the results of studies aiming at assessing representative nutrient values for some Swedish foods and dishes, and at providing measures of the uncertainties for such values. Furthermore, the variations in nutrients due to different sources are investigated. The variations due to chemical determination are assessed for results of analyses of standard samples at different laboratories. Variations due to preparation of dishes are assessed by food preparations under different conditions more or less well controlled. Variations in nutrient content due to sampling of food items from different regions, seasons, and food distribution chains are studied in 15 foods. In the studies, analyses of moisture, ash, nitrogen, fat, calcium, phosphorus, iron, sodium, and potassium are performed, and a reference material suitable for the quality control of such analyses is produced. The produced material, a canned meat homogenate, displayed variations in nutrient content small enough to qualify as a reference material. It was very well received among laboratories performing nutrient analyses and a new material was produced. The interlaboratory standard deviations of results from nutrient analysis were linearly related to the concentrations of the different analytes, except for moisture where a curved relation to concentration seemed to be more relevant. Larger variations were observed for fat in dry materials than for moist materials. For foods sampled in a representative way, the main part of the total variation was due to random effects present when making a purchase. A smaller part could be attributed to systematic factors like season and geographic region. The variations due to preparation of dishes were 4-10% relative standard deviation (RSD%) for standardised preparations in an experimental kitchen, and were more than two times larger, 10-27%, for semi-optional preparations in families following recipies. These figures were compared with results from a third investigation where families had prepared dishes in a totally optional way (20-46% RSD).
显示更多 [+] 显示较少 [-]