We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Department of Mathematical Sciences

Academic Staff

Publication details for Peter Craig

Barcelo Culleres, D, Black, R, Boesten, J, Boobis, A, Hardy, A, Hart, A, Koepp, H, Luttik, R, Machera, K, Maroni, M, McGregor, D, Meyer, O, Moretto, A, Papadopoulou-Mourkidou, E, Petzinger, E, Savolainen, K, Schaeffer, A, Stenström, J, Steurbaut, W, Tsipi-Stefanitsi, D, Vleminckx, C, Ambrus A,, Craig, Peter, Suszter, G & Walters, E (2005). Opinion of the Scientific Panel on Plant health, Plant protection products and their Residues on a request from Commission related to the appropriate variability factor(s) to be used for acute dietary exposure assessment of pesticide residues in fruit and vegetables. Parma, Italy, European Food Safety Authority.

Author(s) from Durham


In a batch of food items previously treated with a pesticide, the residue of the pesticide
remaining on/in single food items at the time of consumption varies between items,
due to a variety of factors. So there is a distribution of residues, with some items
containing more pesticide than others. It is important to take account of this variation
when assessing the risk to consumers from acute dietary exposure to pesticides in
medium and large-sized food items (e.g. apples or melons). Therefore, international
assessment procedures are based on the 97.5th percentile of the distribution of
residues; i.e. the level that is exceeded by 2.5% of residues in food items (i.e. 1 in 40).
This residue level is not measured directly, but estimated by measuring the
concentration in a small batch of items and multiplying it by a “variability factor” to
estimate the 97.5th percentile residue.
Recently, it has been proposed that a default value of 3 should generally be used for
the variability factor, replacing a range of default values for different commodities. The
Commission has asked the PPR Panel to advise on the scientific basis for choosing a
single default value for the variability factor.
The PPR Panel examined the range of variability factors from existing studies where
residues were measured separately in individual food items. The PPR Panel excluded
from this analysis studies for which the variability factor could not be estimated
reliably, i.e. studies with less than 50 items, or where the result was strongly affected
by pesticide residues below the level that could be quantified.
On average, variability factors estimated from samples collected in the marketplace
were higher than those from samples obtained in experimental studies (supervised
trials). The PPR Panel therefore recommends that consideration be given to using
different variability factors when doing exposure assessments with data from these two
types of study.
The average variability factor for supervised trials was 2.8. However, the variability
factor is itself variable, and the Commission may wish to consider this when choosing
an appropriate default value for use in dietary exposure assessments. To assist in this,
the PPR Panel provides tables presenting a range of statistics. For example, it is
estimated that the variability factors for supervised trials will exceed the proposed
default value of 3 in 34% of cases, whereas the previous default value of 7 for mediumsized
food items will be exceeded in 0.2% of cases. Similarly, the variability factors for
market surveys averaged 3.6, and will exceed 3 in about 65% of cases and 7 in about
1% of cases.
The data analysed by the PPR Panel related mostly to medium-sized commodities
(between 25 and 250 g, e.g. apples). The PPR Panel concluded there was insufficient
evidence to support a real difference between variability factors for medium and largesized
commodities, and therefore considered its results applicable to both types.
However, the PPR Panel recommends that this should be re-examined when further
data on large-sized commodities become available.
The results are affected by a number of uncertainties. The PPR Panel calculated
confidence intervals to indicate the degree of uncertainty due to limitations in the
amounts of data available for the analysis. Other uncertainties, e.g. in extrapolating
variability factors between pesticides and between crops, were considered qualitatively.
Finally, the PPR Panel noted that the assessment of acute risks from dietary exposure
uses conservative assumptions for portion size and the mean residue concentration as
well as the variability factor. The combined effect of these conservative assumptions on
the overall level of consumer protection may warrant further consideration.