Concentrations

In the webpage "Concentrations" one is requested to provide one concentration or a series of concentrations of the chemical substance in the selected foods. There are three different ways to do that as shown in the figure below.

Single value   The single value is the method where only one number for the concentration is to be given, e.g the maximum for a worst case evaluation. That number is used for the calculation of the intake by all consumers for all days. In the figure shown here there are 116 subjects studied for 2 days. Then the number of combinations will be 232 (116 subjects × 2 consumption days). The unit for the concentration and a correction for composition and processing are to be set in the fields above. The calculations are done after clicking [Use single value].

Series of concentrations    The second way is to use a file of concentrations as input for all consumers and all days. This file should be a text file; its extension should be '.txt'; each row should contain one value without empty lines between the values as an empty line is considered to be the last line. The number of lines is unlimited. Use decimal points, not commas; "n.d." or "#N/A" to represent non detectable concentrations is not allowed. As an example a series of cadmium concentrations in chocolate (in microgram per kg) is included. For each calculation the tool selects a number at random from the series. The unit for the concentrations and correction for composition and processing are to be set in the first block. These are then used for all calculations. Proceed by clicking [Use this file].

Monte Carlo    For a Monte Carlo calculation two inputs are expected. First a series of concentrations. These are to be provided as a text file as described in the previous section. The second input is the number of repetitions. Now the tool will choose a subject and a consumption day from the database and a concentration from the file, all at random, and will repeat this as many times as the number of repetitions. The unit for the concentrations and a correction for composition and processing are to be set in the first block. These are then used for all calculations. Proceed by clicking [Monte Carlo]. So, the first two methods in the example calculate 116 subjects times 2 intakes, whereas 1000 repetitions will result in a series of 1000 intake values which will contain duplicate values when both subject, day, and concentration are selected more than once at the same time.

GIFTT stores consumption and intake data of individual consumers in a csv file. That file can be downloaded for additional analysis of the results, for example in spreadsheets or statistical packages such as R and SAS. Selection of the Monte Carlo calculation with a large number of repetitions (e.g 50,000 or 100,000) will lead to large csv files. Such files are hard to be converted in a spreadsheet or statistical program because of its size. It is then helpful not to include zero values in the csv file. If 12% of all subjects are consumers of the food of interest while one is asking for 100,000 repetitions, then 88,000 rows of the csv file will contain zeros for consumption and intake. By selection of not to include non-consumers (Include non-consumers in csv file: No) the csv file will contain only those 12,000 rows of all exposed subjects only. That will make import of the csv file in other computer-applications more feasible.