H. van der Voet

Cumulative risk assessment of the exposure to organophosphorus and carbamate insecticides in the dutch diet

We report the acute cumulative exposure to organophosphorus insecticides (OPs) and carbamates in the Dutch population and young children (1-6 years) via the diet. Residue data were derived from Dutch monitoring programmes performed during 2003-2005, and food consumption levels from the Dutch National Food Consumption Survey 1997/1998. The relative potency factor (RPF) approach was used to cumulate the exposure to OPs and carbamates using acephate and oxamyl as index compound respectively. The exposure was estimated using the probabilistic approach, including unit variability and processing effects. We demonstrate that about 3% of the composite samples analysed for OPs and 0.2% for carbamates contain combinations of these pesticides. The P99.9 of exposure to OPs and carbamates in the total Dutch population equals 23 and 0.64microg/kg BW/d respectively. For young children the corresponding exposure levels are 57 and 1.47microg/kg BW/d. When comparing the P99.9 of exposure with the ARfD, 50 and 9microg/kg BW/d for acephate and oxamyl respectively, there is only a possible health risk for young children. Spinach contributed most to the exposure to OPs in both age groups, followed by orange and mandarin. For carbamates apple (sauce) was the main product determining the exposure.

An integrated probabilistic framework for cumulative risk assessment of common mechanism chemicals in food: an example with organophosphorus pesticides.

This paper presents a framework for integrated probabilistic risk assessment of chemicals in the diet which accounts for the possibility of cumulative exposure to chemicals with a common mechanism of action. Variability between individuals in the population with respect to food consumption, concentrations of chemicals in the consumed foods, food processing habits and sensitivity towards the chemicals is addressed by Monte Carlo simulations. A large number of individuals are simulated, for which the individual exposure (iEXP), the individual critical effect dose (iCED) and the ratio between these values (the individual margin of exposure, iMoE) are calculated by drawing random values for all variable parameters from databases or specified distributions. This results in a population distribution of the iMoE, and the fraction of this distribution below 1 indicates the fraction of the population that may be at risk. Uncertainty in the assessment is treated as a separate dimension by repeating the Monte Carlo simulations many times, each time drawing random values for all uncertain parameters. In this framework, the cumulative exposure to common mechanism chemicals is addressed by incorporation of the relative potency factor (RPF) approach. The framework is demonstrated by the cumulative risk assessment of organophosphorus pesticides (OPs). By going through this example, the various choices and assumptions underlying the cumulative risk assessment are made explicit. The problems faced and the solutions chosen may be more generic than the present example with OPs. This demonstration may help to familiarize risk assessors and risk managers with the somewhat more complex output of probabilistic risk assessment.

Introduction to the Monte Carlo project and the approach to the validation of probabilistic models of dietary exposure to selected food chemicals

The Monte Carlo project was established to allow an international collaborative effort to define conceptual models for food chemical and nutrient exposure, to define and validate the software code to govern these models, to provide new or reconstructed databases for validation studies, and to use the new software code to complete validation modelling. Models were considered valid when they provided exposure estimates (ea) that could be shown not to underestimate the true exposure (eb), but at the same time are more realistic than the currently used conservative estimates (ec). Thus, validation required eb⩽ea<ec. In the case of pesticides, validation involved the collection of duplicate diets from 500 infants for pesticide analysis. In the case of intense sweeteners, a new consumption dataset was created among prescreened high consumers of intense sweeteners by recording, at brand level, all foods and beverages ingested over 12 days. In the case of nutrients and additives, existing databases were modified to minimize uncertainty over the model parameters. In most instances, it was possible to generate probabilistic models that fulfilled the validation criteria.

Validation of a probabilistic model of dietary exposure to selected pesticides in Dutch infants

A probabilistic model for dietary exposure to pesticides was validated. For this, we evaluated the agreement of dietary exposure to six pesticides as estimated with the model with exposures measured in duplicate diet samples (=‘real intake’) and those calculated with the point estimate. To calculate the exposure with the model and point estimate, consumption data of the duplicate diet survey and pesticide residue measurements from Dutch monitoring programmes in 2000 and 2001 were used. The model was considered validated when the outcome was both higher than the real intake and lower than the point estimate. Results showed that exposures estimated with the model were closer to the real intake than those of the point estimate, and that the model outcome was lower than the point estimate. Furthermore, it was shown that the probabilistic approach can address the exposure to a pesticide via the consumption of different food products, while the point estimate only estimates the exposure through the consumption of one product. The model validated is a valuable asset when estimating the dietary exposure to pesticides in both the authorization of new pesticides and the evaluation of exposures using monitoring data.

A model for probabilistic health impact assessment of exposure to food chemicals.

A statistical model is presented extending the integrated probabilistic risk assessment (IPRA) model of van der Voet and Slob [van der Voet, H., Slob, W., 2007. Integration of probabilistic exposure assessment and probabilistic hazard characterisation. Risk Analysis, 27, 351-371]. The aim is to characterise the health impact due to one or more chemicals present in food causing one or more health effects. For chemicals with hardly any measurable safety problems we propose health impact characterisation by margins of exposure. In this probabilistic model not one margin of exposure is calculated, but rather a distribution of individual margins of exposure (IMoE) which allows quantifying the health impact for small parts of the population. A simple bar chart is proposed to represent the IMoE distribution and a lower bound (IMoEL) quantifies uncertainties in this distribution. It is described how IMoE distributions can be combined for dose-additive compounds and for different health effects. Health impact assessment critically depends on a subjective valuation of the health impact of a given health effect, and possibilities to implement this health impact valuation step are discussed. Examples show the possibilities of health impact characterisation and of integrating IMoE distributions. The paper also includes new proposals for modelling variable and uncertain factors describing food processing effects and intraspecies variation in sensitivity.

Inter-laboratory, time, and fitness-for-purpose aspects of effective validation

Abstract This paper reviews current approaches to validation of analytical chemical methods. It identifies some shortcomings of existing validation schemes such as insufficient coverage of variability in space or time, and mismatches between validation criteria and intended use of the method, e.g. the use in regulatory control. A general statistical modelling approach for combining different aspects of validation is recommended, and illustrated with an example. This type of modelling, based on components of variation, is advocated as the basis for the development of new statistically underpinned validation schemes which integrate current validation and quality assurance activities.

Prevalence of Salmonella in the broiler supply chain in the Netherlands

This article presents detailed information on Salmonella prevalence throughout the broiler supply chain in The Netherlands, based on results from a national monitoring program. Data were collected during the period 2002 through 2005 and from six sampling points in the chain, covering hatchery up to and including processing. Trends in Salmonella prevalence over years and seasons were analyzed as well as the effect of slaughterhouse capacity on these trends. In addition, correlations between the occurrence of Salmonella at the various sampling points were calculated. The results showed a decreasing trend of Salmonella prevalence from 2002 through 2005 at all sampling points. A seasonal effect on the occurrence of Salmonella was found at the broiler farm, with a higher prevalence during the third and fourth quarter of the year (July through December). The higher the capacity of the slaughterhouse, the lower Salmonella prevalence on arrival at the slaughterhouse and the higher the prevalence at the end of slaughter and the end of processing. The detailed insights obtained in this study could be used to focus future field and experimental research on the prevention and control of Salmonella in the broiler supply chain. Results presented could also be used in risk assessment studies.

Can current dietary exposure models handle aggregated intake from different foods? A simulation study for the case of two foods.

Current dietary exposure models provide estimates of long-term intake distributions using short-term food consumption survey data, by statistically modeling the aggregated intakes from different foods consumed on the same day for each participant of the survey. Food consumption behaviour in a population may, however, show all sorts of correlations which are not modelled in these exposure models. We developed a simulation model describing a hypothetical population of consumers, assuming various types of correlation between two foods. Using this simulation model we found that the impact of the correlations in many cases is limited, but in particular circumstances it can be substantial, depending on the properties of the marginal distributions. Further, we found that the usual approach of first aggregating the observed intakes over foods, and then applying the statistical exposure models to the total daily intakes may lead to deviating results, even when all correlations are assumed to be zero. The approach of analyzing the intakes from the separate foods, and then aggregating the results from the statistical model applied to each food performed much better. Our results illustrate that the simulation model can be used for validating dietary exposure models, and for indicating how exposure models may be improved.

Increased efficacy for in-house validation of real-time PCR GMO detection methods

To improve the efficacy of the in-house validation of GMO detection methods (DNA isolation and real-time PCR, polymerase chain reaction), a study was performed to gain insight in the contribution of the different steps of the GMO detection method to the repeatability and in-house reproducibility. In the present study, 19 methods for (GM) soy, maize canola and potato were validated in-house of which 14 on the basis of an 8-day validation scheme using eight different samples and five on the basis of a more concise validation protocol. In this way, data was obtained with respect to the detection limit, accuracy and precision. Also, decision limits were calculated for declaring non-conformance (>0.9%) with 95% reliability. In order to estimate the contribution of the different steps in the GMO analysis to the total variation variance components were estimated using REML (residual maximum likelihood method). From these components, relative standard deviations for repeatability and reproducibility (RSDr and RSDR) were calculated. The results showed that not only the PCR reaction but also the factors ‘DNA isolation’ and ‘PCR day’ are important factors for the total variance and should therefore be included in the in-house validation. It is proposed to use a statistical model to estimate these factors from a large dataset of initial validations so that for similar GMO methods in the future, only the PCR step needs to be validated. The resulting data are discussed in the light of agreed European criteria for qualified GMO detection methods.

Probabilistic modelling of exposure doses and implications for health risk characterization: Glycoalkaloids from potatoes

Potatoes are a source of glycoalkaloids (GAs) represented primarily by alpha-solanine and alpha-chaconine (about 95%). Content of GAs in tubers is usually 10-100 mg/kg and maximum levels do not exceed 200 mg/kg. GAs can be hazardous for human health. Poisoning involve gastrointestinal ailments and neurological symptoms. A single intake of >1-3 mg/kg b.w. is considered a critical effect dose (CED). Probabilistic modelling of acute and chronic (usual) exposure to GAs was performed in the Czech Republic, Sweden and The Netherlands. National databases on individual consumption of foods, data on concentration of GAs in tubers (439 Czech and Swedish results) and processing factors were used for modelling. Results concluded that potatoes currently available at the European market may lead to acute intakes >1 mg GAs/kg b.w./day for upper tail of the intake distribution (0.01% of population) in all three countries. 50 mg GAs/kg raw unpeeled tubers ensures that at least 99.99% of the population does not exceed the CED. Estimated chronic (usual) intake in participating countries was 0.25, 0.29 and 0.56 mg/kg b.w./day (97.5% upper confidence limit). It remains unclear if the incidence of GAs poisoning is underreported or if assumptions are the worst case for extremely sensitive persons.