1.2.2 Chronic exposure

In the case of chronic dietary exposure the same approaches can be used as for the acute dietary exposure assessments. However, the parameters used differ.

Point estimate

This technique is also used worldwide and the same formula as for acute exposure is applied:

(consumption level x compound level)  / body weight

Opposed to the acute exposure, in the chronic exposure assessment a mean consumption level is used, together with a mean, median or MRL (maximum residue level) residue level of the compound. The end result, as in the acute exposure, is one level of exposure and no information is given on how many people have such an exposure or have exceeded it. Total intake of the compound, when present in more foods, can be obtained by cumulating the intakes from all foods containing the compound.

Deterministic approach

For this technique, as with acute exposure, a single compound level (mean, median, MRL) and a distribution of consumption levels are used. To obtain a distribution of overall exposure levels, the exposure levels are summed over foods In this approach a distribution of mean daily (observed) exposure levels are obtained. No attempt is made to extrapolate these estimates to chronic exposure levels.

Probabilistic modelling

This technique uses a single mean (or median) compound level and a distribution of daily consumption levels, as in the deterministic approach. However, to extrapolate this distribution of daily mean exposure levels to a distribution of chronic (or long-term) exposure levels statistical tools are used, such as the method of Slob {see references}. For the chronic exposure the interest is in the variation between the individuals. The variation between the different food recording days for the same individual is not relevant. Therefore, a distribution of the chronic exposure will be considerably narrower than the distribution of daily mean exposure as obtained with the deterministic approach. Mean daily exposure levels tend to over-estimate long-term exposure levels.

To use this method information on intra- and inter individual variation in a population is needed. This can be obtained from consumption data with information on daily eating habits for each person of at least 2, preferably, non-consecutive days.