GIFTT

The GIFT Tool is to be used to calculate intake of chemical substances from foods for individual consumers. This is done using consumption data that are collected in various food consumption surveys from different countries in the world. These data can be found and be downloaded from the FAO/WHO GIFT website. At of March 2026 the database covers 67 surveys with only a limited number of European countries (as unfortunately most surveys in Europe are still kept confidential). Not all surveys in GIFT are yet available in the tool, but in the forthcoming period more suveys will be added. Additional available databases with consumption figures of the Netherlands and Cyprus were added to the database in the Foodsafetyportal. The XI and ImproRisk tool are therefore deleted from the Portal and must be now addressed through GIFT.

An important topic in the calculation of exposure of individual consumers is the possibility to perform "Monte Carlo" calculations. With this it will be possible to find the common variation of the exposure, and to define whether or not most of the variation comes from the variation of consumption quantities, of from the concentrations of the chemical substances in foods, or other variables such as consumers' body weights.

Monte Carlo

The calculation of the dietary intake is a "point estimate", based on one number for the consumption and one for the concentration in the food, for an average or extreme consumer. All values come from a range of data, and there are no fixed rules about what number to select. The tendency is to use summarized consumption quantities, and it is oft advised to select the higher concentrations (95th or 99th percentile) for acute exposure, with the most sensitive consumer (low body weight, most often children). So, using higher percentiles gives higher intake numbers but with a lower probability, and a higher chance of rejection of foods that are actually safe. To overcome the differences between a point estimate and the exposure of a population one can use "probabilistic modelling". It is based on a calculation where all distributions are multiplied with each other, instead of one value taken from each distribution. For such calculations a tool is needed that reads the consumption quantities of all consumers and the concentrations of all samples. Then the computer selects a number randomly from the data, calculates the intake, giving one result. Then it selects another one, and so on, leading to a series of intakes. The output of a Monte Carlo exposure calculation is a series of numbers of intakes of individual consumers. These are is well suited for statistical analysis, e.g. to determine various percentiles and their probability.