## What is exposure

Exposure to a substance found in foodstuffs, irrespective of source, is derived as follows:

• For each individual

- Exposure for food item = concentration of the substance in a food item x weight of item consumed.

- Exposure for meal = sum of exposure of all items consumed during that meal.

- Exposure over lifetime = sum of exposure for all meals.

• For the population

Exposure for population = distribution of exposure for every individual obtained by repeating the above three steps for the entire population.

The exposure intake for any population is a sum of the products of the concentration (c) in mg/kg or mg/kg of the migrant within the food item eaten and the weight (w) in kilograms of that item, and is expressed as (Holmes et al. 2005)

Here, DDjk is the daily dose for any individual j on day k consuming up to n(k) items on that day. Wj is the weight of the individual j and Cjkl is the concentration of the migrant in the food item l, whilst Wjkl denotes the weight of item l on day k eaten by individual j.

A distribution curve similar to that shown in Fig. 6.1 can be obtained. From the 'y' axis it is possible to select any percentile consumer and 'read off' the exposure on the 'x' axis. If all of these data existed with no uncertainty, then there would be no need to estimate exposure. Having obtained an exposure curve, it is then necessary to decide what percentile value should be used for the consequent risk assessment. The percentile to use for the 'protection' of any population to any risk is a political and not a scientific decision. The regulators and politicians have to be realistic as 100% of the population can never be protected all of the time. The considered opinion of many is that the diet of the consistent 100th percentile consumer would result in serious health problems due to nutritional factors long before any effects due to migrants from packaging would be a factor. The definition of the high consumer varies, but for food contact materials, either the 90th, 95th or 97.5th percentile is typically used for the high consumer (Kroes et al. 2002) rather than the maximum value.

This does not mean that 10, 5 or 2.5% of the population are unprotected, but as a consequence of the derivation of the percentile, a high consumer of one food type is most unlikely to be a high consumer of another and certainly not for a lifetime. Furthermore, the maximum level consumer is unlikely to maintain this consumption over long periods of time and therefore is not representative of high level intakes in relation to chronic exposure (Benford and Tennant 1997). The use of the 95th percentile has been considered as a

95th percentile consumer has 4.5 mg/day

95th percentile consumer has 4.5 mg/day

Exposure mg/day

Fig. 6.1 Cumulative exposure curve for a population showing an exposure ('x' axis) for a given percentile ('y' axis) (Holmes et al. 2005).

good compromise (Chambolle 1999) between a high level of protection for the consumer and the uncertainty (or precision) in the intake estimates. However, as a consequence there is correlation between the percentile selected and the necessary precision and the sample size. For a given sample size the nearer one gets to the 100th percentile the fewer the consumers. It is now possible to statistically treat scenarios for high consumers with small sample sizes but the uncertainty of any estimate increases.

Whilst it is necessary to consider all sources for a substance to be found in foodstuffs for a full exposure assessment, this chapter considers only contamination of food from its packaging. For the presence of substances which could be in packaging as well as in the raw food (however few), any contamination from food processing equipment and any migration from cookware or cooking utensils in the home, once the food has been emptied from its packaging are outside the scope of this chapter. Today, the exposure to a substance migrating from packaging into food cannot be determined exactly, as consumption data with the packaging of the foodstuff consumed do not exist for the whole of the European population. Indeed, very limited data exist for the EU and in some cases even the surveys of the foodstuffs are barely adequate, but their prime purpose is for nutrition.

0 0