Total consumer exposure to polybrominated diphenyl ethers in North America and Europe

Polybrominated diphenyl ethers (PBDEs) have been widely used as flame retardants in textiles, polyurethane foams, and plastics. PBDEs exert toxic effects in various organisms, including humans, and are ubiquitous in the outdoor and indoor environment. Here we estimate total daily PBDE doses received by consumers in North America and Europe, along with the most important pathways and congeners, and derive PBDE elimination half-lives for chronic exposure. We estimate distributions for all parameters (PBDE concentrations in exposure media, food consumption rates, etc.) and conduct a probabilistic exposure assessment. We find that Americans are exposed the most, likely due to stricter fire regulations, followed by consumers from the UK and Continental Europe. In the central quantiles of the exposure distributions derived, food is the dominant pathway; in the upper quantiles either food or oral and dermal exposure to dust. This reflects the lipophilic and persistent nature of PBDEs and their use in products for indoor-use. Median elimination half-lives are in a range of 1-3 years except for BDE-153 with about seven years and BDE-209 with 4-7 days.     

A statistical method to determine whether micro-organisms are randomly distributed in a food matrix,applied to coliforms and Escherichia coli O157 in minced beef

A method is presented to determine whether micro-organisms are distributed randomly or not in a batch of food. This is important information for quantitative microbiological risk assessment (QMRA). If micro-organisms are distributed randomly, this variation in numbers of micro-organisms between samples from a batch of food theoretically agrees with a Poisson distribution, with a fixed value for parameter λ. Other sources for variation (such as clustering of micro-organisms or experimental error) can lead to an increase of variation of the data, which can be described by a Poisson distribution with a Gamma distributed λ (Poisson(Gamma)). The Poisson and Poisson(Gamma) distributions were used to describe the effect of grinding on the variation of endogenous flora of coliforms and artificial, clustered contamination of Escherichia coli O157 in minced beef. In addition, the Poisson and Poisson(Gamma) distributions were compared to the Lognormal distribution. Coliforms were enumerated in 25 samples in the starting material and in the mince produced after grinding once, twice and three times. E. coli O157 was spiked into the starting material in 1 or 10 clusters, and enumerated in 25 samples after grinding once and twice. Coliforms were not randomly distributed in the starting material. Grinding resulted in random and non-random distributed coliforms and E. coli O157, respectively. Generally, the Lognormal and Poisson(Gamma) distribution fitted equally well to the data, but only the Poisson(Gamma) distribution is useful to determine whether micro-organisms are randomly distributed or not.     

Quantitative dietary exposure assessment of the Catalonian population (Spain) to the mycotoxin deoxynivalenol

The mycotoxin deoxynivalenol (DON) is one of the most common contaminants of cereals worldwide, and its occurrence has been widely reported in raw foods and foodstuffs, around the European region, including Catalonia. In the present work, a stochastic methodology has been applied to accurately assess the exposure of the Catalonian population (Spain) to DON through food consumption. Raw contamination data was provided by a large survey conducted in this region, in addition to the raw consumption data from a nutritional study specifically designed to assess the dietary intake of the main foodstuffs related to DON contamination for all population age groups. Contamination and consumption data were combined by simulation using an essentially parametric (P-P) method. The P-P method draws sampling values from distribution functions fitted to consumption and contamination data sets. Moreover, to quantify the accuracy and reliability of the statistics estimates, we built the related confidence intervals using a pseudo-parametric bootstrap method. Considering the results drawn from the P-P simulation method, the Catalonian population should be expected to be exposed at moderated levels of deoxynivalenol, the infants and individuals with ethnic dietary patterns being the most exposed population groups.     

Quantitative dietary exposure assessment of the Catalonian population (Spain) to the mycotoxin deoxynivalenol

The mycotoxin deoxynivalenol (DON) is one of the most common contaminants of cereals worldwide, and its occurrence has been widely reported in raw foods and foodstuffs, around the European region, including Catalonia. In the present work, a stochastic methodology has been applied to accurately assess the exposure of the Catalonian population (Spain) to DON through food consumption. Raw contamination data was provided by a large survey conducted in this region, in addition to the raw consumption data from a nutritional study specifically designed to assess the dietary intake of the main foodstuffs related to DON contamination for all population age groups. Contamination and consumption data were combined by simulation using an essentially parametric (P-P) method. The P-P method draws sampling values from distribution functions fitted to consumption and contamination data sets. Moreover, to quantify the accuracy and reliability of the statistics estimates, we built the related confidence intervals using a pseudo-parametric bootstrap method. Considering the results drawn from the P-P simulation method, the Catalonian population should be expected to be exposed at moderated levels of deoxynivalenol, the infants and individuals with ethnic dietary patterns being the most exposed population groups     

Statistical method to assess usual dietary intakes in the European population

Food consumption data are a key element of EFSA’s risk assessment activities, forming the basis of dietary exposure assessment at the European level. In 2011, EFSA released the Comprehensive European Food Consumption Database, gathering consumption data from 34 national surveys representing 66,492 individuals from 22 European Union member states. Due to the different methodologies used, national survey data cannot be combined to generate European estimates of dietary exposure. This study was executed to assess how existing consumption data and the representativeness of dietary exposure and risk estimates at the European Union level can be improved by developing a ‘Compiled European Food Consumption Database’. To create the database, the usual intake distributions of 589 food items representing the total diet were estimated for 36 clusters composed of subjects belonging to the same age class, gender and having a similar diet. An adapted form of the National Cancer Institute (NCI) method was used for this, with a number of important modifications. Season, body weight and whether or not the food was consumed at the weekend were used to predict the probability of consumption. A gamma distribution was found to be more suitable for modelling the distribution of food amounts in the different food groups instead of a normal distribution. These distributions were combined with food correlation matrices according to the Iman–Conover method in order to simulate 28 days of consumption for 40,000 simulated individuals. The simulated data were validated by comparing the consumption statistics of the simulated individuals and food groups with the same statistics estimated from the Comprehensive Database. The opportunities and limitations of using the simulated database for exposure assessments are described.     

Dietary exposure to chemical migrants from food contact materials: A probabilistic approach

A two-dimensional probabilistic model has been developed to estimate the short-term dietary exposure of UK consumers to migrants from food packaging materials. The current EU approach uses a default scenario of assuming that all individuals are 60 kg weight and consume 1 kg of food packaged in the material of interest per day. Using four UK National Dietary and Nutrition Surveys comprising 4-7 day dietary records for different age groups and survey years, a sample representative of the UK population has been obtained consuming around 4200 different food items. Each survey provides records for around 2000 individuals and supplies detailed information on the consumption of food and data on sex, height and socio-economic status which may be used to analyse the exposure of selected groups within the community. As a result we are able to address the variation in consumption of food amongst individuals, and account for actual body weights providing a more accurate representation of the 'true' exposure. The migrants bisphenol A diglycidyl ether (BADGE), di-2-ethylhexyl adipate (DEHA) and styrene were considered as specimen compounds although the methodology employed has the flexibility to adapt to other migrants and packaging types and indeed other food contaminants. Exposure for each individual is estimated by calculating and summing the individual exposure from each item in their diet, and is repeated for all individuals in each survey to produce a distribution of exposures for the population. The packaging type of each food item is assigned by utilizing known packaging types from the database or, by sampling from a distribution based upon market share information. The parameters contributing towards the exposure from a packaged dietary item are migrant concentration and item weight. Distributions are used to represent the inherent variation and uncertainty affecting these parameters. Where data on concentrations for a particular type of food are lacking, expert judgement is used to extrapolate from available data for other food types. The model can also be run using only migration data for food simulants. In this case, concentrations expected for each of the food items are assigned based on the data for the relevant food simulant. The primary outputs of the model are distributions of estimated daily intakes for the selected population. Each distribution gives the variation across the population subject to the uncertain parameters sampled in that iteration of the model. Analysing the ensemble of distributions allows us to obtain the confidence limits around estimates for percentiles due to the uncertainties. The probabilistic approach allows sensitivity analysis to evaluate the relative importance of the input parameters and places confidence bounds on the outputs to show the effect of the uncertainties and the contribution of each food type toward the overall exposure.     

Assessment of food intake input distributions for use in probabilistic exposure assessments of food additives

A key component of a food chemical exposure assessment using probabilistic analysis is the selection of the most appropriate input distribution to represent exposure variables. The study explored the type of parametric distribution that could be used to model variability in food consumption data likely to be included in a probabilistic exposure assessment of food additives. The goodness-of-fit of a range of continuous distributions to observed data of 22 food categories expressed as average daily intakes among consumers from the North-South Ireland Food Consumption Survey was assessed using the BestFit® distribution fitting program. The lognormal distribution was most commonly accepted as a plausible parametric distribution to represent food consumption data when food intakes were expressed as absolute intakes (16/22 foods) and as intakes per kg body weight (18/22 foods). Results from goodness-of-fit tests were accompanied by lognormal probability plots for a number of food categories. The influence on food additive intake of using a lognormal distribution to model food consumption input data was assessed by comparing modelled intake estimates with observed intakes. Results from the present study advise some level of caution about the use of a lognormal distribution as a mode of input for food consumption data in probabilistic food additive exposure assessments and the results highlight the need for further research in this area.     

Hierarchical Bayesian analysis of censored microbiological contamination data for use in risk assessment and mitigation

Microbiological contamination data often is censored because of the presence of non-detects or because measurement outcomes are known only to be smaller than, greater than, or between certain boundary values imposed by the laboratory procedures. Therefore, it is not straightforward to fit distributions that summarize contamination data for use in quantitative microbiological risk assessment, especially when variability and uncertainty are to be characterized separately. In this paper, distributions are fit using Bayesian analysis, and results are compared to results obtained with a methodology based on maximum likelihood estimation and the non-parametric bootstrap method. The Bayesian model is also extended hierarchically to estimate the effects of the individual elements of a covariate such as, for example, on a national level, the food processing company where the analyzed food samples were processed, or, on an international level, the geographical origin of contamination data. Including this extra information allows a risk assessor to differentiate between several scenario’s and increase the specificity of the estimate of risk of illness, or compare different scenario’s to each other. Furthermore, inference is made on the predictive importance of several different covariates while taking into account uncertainty, allowing to indicate which covariates are influential factors determining contamination.     

Biomarkers of exposure to fumonisin mycotoxins: A review

The investigation of adverse health effects associated with fungal mycotoxins requires the measurement of human exposure. Most frequently, this exposure is estimated from contamination levels of raw foodstuffs, which are the primary source of toxin exposure, and data on food consumption patterns. However, variations in food preparation methods, food intake, contamination level, intestinal absorption, toxin distribution and excretion lead to individual variations in toxin exposure that are more readily measured with a biomarker. Fumonisin biomarkers have been sought in the measurement of levels of the toxin in physiological samples such as serum, urine, faeces, hair and nails. However, due to the low bioavailability of fumonisin, these samples pose a variety of analytical challenges and also still require validation as biomarkers. The most widely researched fumonisin biomarkers have been those related to the disruption of de novo sphingolipid biosynthesis, namely elevated levels of the sphingoid base, sphinganine, or of its ratio with sphingosine. Elevation of these parameters in humans would potentially provide a biomarker of biochemical effect. A number of investigations into the possible elevation of sphinganine (or its ratio with sphingosine) in human blood and urine have generally failed to correlate with estimates of fumonisin exposure. The sphingoid bases occur naturally in human blood and urine such that their levels have normal ranges, which can be influenced by dietary factors other than fumonisin ingestion. The lower exposures from human diets, as compared with doses in experimental animals, have made detection of changes in these sphingoid biomarkers problematic.     

Pesticides on household surfaces may influence dietary intake of children

The physical and chemical environment influences children's exposures to pesticides in and around the home. Children's activities, which increase their potential for exposure especially during eating, have been captured in the Children's Dietary Intake Model (CDIM). In addition to the chemical exposure associated with the food itself, this model incorporates excess dietary exposures due to handling of food during consumption. To stochastically evaluate CDIM, distributions of measured, and in some cases estimated, model factors were determined from measurements of permethrin, chlorpyrifos, and diazinon derived from assembled databases and laboratory experiments. Using the distributions of these factors, Monte Carlo simulations were performed to obtain distributions of total dietary intake of pesticides. To target the sources of pesticide contamination that were influencing total dietary intake, each factor was evaluated. We found pesticide surface concentration to be highly influential. By excluding surface concentration, we were also able to determine the influence of the other factors based on the F-statistic. Transfer efficiencies, followed by pesticide residue in consumed foods and amount of food consumed, were the next most influential factors within the model. With these distributions for model inputs, CDIM has the potential to more accurately predict total dietary intake of a contaminant by a child.     

常德市餐饮食品金黄色葡萄球菌检测结果分析

目的 通过对餐饮食品金黄色葡萄球菌的检测, 了解常德市餐饮食品金黄色葡萄球菌污染情况。方法 对常德市330家餐饮企业的13类473批餐饮食品采样, 检测金黄色葡萄球菌。结果 餐饮食品的金黄色葡萄球菌检测合格率为99.2％, 在凉拌菜、非发酵型豆制品、米粉中检出金黄色葡萄球菌, 13类食品的金黄色葡萄球菌检测合格率有显著性差异(X2=23.88, P＜0.05), 2011年～2013年金黄色葡萄球菌检测合格率无显著性差异(X2=3.03, P＞0.05), 6类餐饮业态的金黄色葡萄球菌污染率无显著性差异(X2=6.36, P＞0.05)。结论 常德市餐饮食品中存在金黄色葡萄球菌污染, 需加强对易受金黄色葡萄球菌污染的米粉、凉拌菜等重点品种的监管。     

Dietary exposure and risk assessment to lead of the population of Jiangsu province,China

The paper’s main purpose is to estimate the dietary exposure to lead for the inhabitants of Jiangsu province, China. Lead concentration data were obtained from the national food contamination monitoring programme during 2007–10. Food samples (n = 2077) were collected from 23 food categories in Jiangsu province. Consumption data were derived from Chinese national nutrition and health survey in 2002, which included 3938 inhabitants from 1451 households in Jiangsu province. Concentration data were combined with consumption data to estimate the dietary intake for the inhabitants of 2–6, 7–17 and 18–80 years, respectively. The β-binomial–normal (BBN) model was used to estimate the long-term intake for the population in Jiangsu province. The distribution of individual margin of exposure (IMoE) was introduced to assess the health effect. Uncertainty of IMoE was quantified by Monte Carlo and bootstrap methods. The mean levels of dietary exposure to lead were estimated at 3.019 µg kg^?1 bw day^?1 for children aged 2–6 years, 2.104 µg kg^?1 bw day^?1 for teenagers aged 7–17 years, and 1.601 µg kg^?1 bw day^?1 for adults aged 18–80 years. The mean intakes for the urban and rural populations were 1.494 and 1.822 µg kg^?1 bw day^?1, respectively. From the 25th to 99.9th percentiles, IMoE was 0.125–2.057 for 2–6 years and 0.473–7.998 for 18–80 years, respectively. The distribution of IMoE could indicate a public health concern on lead for the Chinese population in Jiangsu. Control measures should be taken to reduce lead exposure in Jiangsu province.     

Modelling homogeneous and heterogeneous microbial contaminations in a powdered food product

The actual physical distribution of microorganisms within a batch of food influences quantification of microorganisms in the batch, resulting from sampling and enumeration by microbiological tests. Quantification may be most accurate for batches in which microorganisms are distributed homogeneously. However, when the distribution is non-homogeneous, quantification may result in an under-, or overestimation. In the case of pathogens being non-homogeneously distributed, this heterogeneity will impact on public health. Enumeration data are commonly modelled by the Lognormal distribution. Although the Lognormal distribution can model heterogeneity, it does not allow for complete absence of microorganisms. Studies that validate the appropriateness of using Lognormal or other statistical distributions are scarce. This study systematically investigated laboratory and industrial scale batches of powdered infant formula, modelled the enumeration data using a range of statistical distributions, and assessed the appropriateness of individual models. For laboratory scale experiments, batches of milk powder were contaminated by distributing similar numbers of cells of Cronobacter sakazakii either homogeneously throughout a batch of milk powder or by distributing the cells in a localised part of the batch. Each batch was then systematically sampled and the distribution determined by enumerating the samples. By also enumerating the remainder of the batch, a balance could be made of the total number of microorganisms added and of the number retrieved from a batch. Discrete, as well as continuous statistical distributions, were fitted to enumeration data and the parameters estimated by Maximum Likelihood. The data were fitted both as censored and uncensored data. Enumeration data obtained for an industrial batch of powdered infant formula were investigated in this way as well. It was found that Normal, Poisson and Zero-Inflated Poisson distributions fitted the data sets very poorly. In case of homogeneous contamination, there was not a notable difference between the ability of Negative Binomial, Poisson-Lognormal, Weibull, Gamma, and Lognormal distributions to model the data. Overall, either the Negative Binomial distribution or the Poisson-Lognormal distribution fitted the data best in the 10 batches studied, especially when part of a data set contained zeros and/or the numbers were low. The Negative Binomial fitted the laboratory batches best and the Poisson-Lognormal fitted the industrial batch best.     

The Gamma-Poisson model as a statistical method to determine if micro-organisms are randomly distributed in a food matrix

The Gamma-Poisson model, i.e., a Poisson distribution where the parameter lambda is Gamma distributed, has been suggested as a statistical method for determining whether or not micro-organisms are randomly distributed in a food matrix. In this study, we analyse the Gamma-Poisson model to explore some of the properties of the Gamma-Poisson model left unexplored by the previous study. The conclusion of our analysis is that the Gamma-Poisson model distinguishes poorly between variation at the Poisson level and the Gamma level. Estimated parameter values from simulated data-sets showed large variation around the true values, even for moderate sample sizes (n=100). Furthermore, at these sample sizes the likelihood ratio is not a good test statistic for discriminating between the Gamma-Poisson distribution and the Poisson distribution. Hence, to determine if data are randomly distributed, i.e., Poisson distributed, the Gamma-Poisson distribution is not a good choice. However, the ratio between variation at the Poisson level and the Gamma level does provide a measure of the amount of overdispersion.     

Current importance of ochratoxin A-producing Aspergillus spp

Ochratoxin A (OA) is receiving attention worldwide because of the hazard it poses to human and animal health. OA contamination of commodities, such as cereals or pork and poultry meat, is well recognized. Nevertheless, there is an increasing number of articles reporting OA contamination in other food commodities, such as coffee, beer, wine, grape juice, and milk, in the last few years. This continuous and increasing exposure to OA that humans experience is reflected in the high incidence of OA in both human blood and milk in several countries. OA was believed to be produced only by Aspergillus ochraceus and closely related species of section Circumdati and by Penicillium verrucosum; however, in the genus Aspergillus, the production of OA has been recently reported by species outside the section Circumdati. Thus, it has been clearly established as a metabolite of different species of the section Nigri, such as Aspergillus niger and Aspergillus carbonarius. OA production ability by Aspergillus spp. is more widespread than previously thought; therefore, there is the possibility that unexpected species can be new sources of this mycotoxin in their natural substrates.     

Simulation of deoxynivalenol intake from wheat consumption in Japan using the Monte Carlo method

The aim of this study was to evaluate the current advisory level in Japan for deoxynivalenol (DON) in foods. To this end, we estimated the intake of DON based on its presence in wheat using a probabilistic computer simulation method. Values for the concentration of DON in wheat were based on those reported in surveys of 638 wheat samples conducted from 2002 to 2004. Data regarding consumption of 108 wheat-based products according to age group were obtained from the 2002 Japan national survey on food consumption. Two data sets on the consumption of wheat-based products and contamination of DON in wheat were analysed using three DON regulatory scenarios: no regulation, 1100 μg kg(-1) and 2000 μg kg(-1). Because consumption distributions contained two peaks for each age category, it was assumed that two log-normal distributions for each age category were needed to achieve a better fit to the distribution models. The results of simulated DON intake using the Monte Carlo method showed that children aged 1-6 years have the highest DON intake. However, the 95th percentile of simulated intake of DON in each age group was below the provisional maximum tolerable daily intake (TDI) of 1 μg kg(-1) body weight using any regulation scenario. The 99th percentile of simulated DON intake in the 1-6-year-old group was greater than TDI at approximately 2 μg kg(-1) body weight. These results suggest that the current dietary intake of DON from wheat consumption does not exert a significant health effect, but we may need to reconsider the current regulation value for the 1-6-year-old age group. In addition, we may need a better method to fit the distribution to the log-normal distribution better.     

Wild mushroom consumption by pickers in the south of Spain: a probabilistic approach

A survey has been made of 300 habitual consumers of the eight species of wild mushrooms most often consumed in the south of Spain. The eight species selected constitute over 95% of the intake of this food in the samples studied. The mean consumption per capita of mushrooms in Spain is of 10.4 kg/year, 8.6 kg of which are consumed during the season, which lasts from between 1 and 3 months. Male pickers from the Huelva province were those who presented the largest intake, their age group being highly influenced by the species. The consumption of each mushroom studied and the total intake were adjusted/fitted to exponential distributions. These distributions could be an effective tool for toxicological or nutritional studies since they permit the evaluation of exposure that makes it possible to calculate the probabilistic risk analysis and the contribution to the reference dietary intake, respectively, for this population group.     

Assessment of dietary exposure of young Chinese children to aluminium residues

The aim of this study was to assess the Al dietary exposure of young Chinese children aged 0–3 years via formulae, complementary foods and wheat-based foods. Al residue data were obtained from the national food contamination monitoring programme from 2013 to 2016, encompassing 13,833 samples of 12 food items with a detection rate of 76.0%. Food consumption data were gathered from the China National Food Consumption Survey conducted in 2015, comprising 20,172 children aged 0–3 years old. The mean dietary exposure to Al for the general population of young Chinese children was estimated at 0.76 mg/kg bw/week, which does not exceed the PTWI. The 97.5th percentile intakes of Al reached 3.42 mg/kg bw/week, more than 1.7 times the PTWI. Wheat-based foods contributed 80.5% of the Al intake for the general population of young Chinese children, while formulae and complementary foods accounted for 19.5% of the total intake. The dietary intake of Al from formulae and complementary foods accounted for 6.0% and 1.6% of PTWI, respectively. These findings suggested that dietary exposure to Al among the general population of young Chinese children was lower than the PTWI and that there are no health concerns related to this level of Al intake. However, more attention should be placed on the health risks associated with Al exposure from wheat-based foods for young consumers with high food consumption in China (97.5th percentile)     

Simulation of deoxynivalenol intake from wheat consumption in Japan using the Monte Carlo method

The aim of this study was to evaluate the current advisory level in Japan for deoxynivalenol (DON) in foods. To this end, we estimated the intake of DON based on its presence in wheat using a probabilistic computer simulation method. Values for the concentration of DON in wheat were based on those reported in surveys of 638 wheat samples conducted from 2002 to 2004. Data regarding consumption of 108 wheat-based products according to age group were obtained from the 2002 Japan national survey on food consumption. Two data sets on the consumption of wheat-based products and contamination of DON in wheat were analysed using three DON regulatory scenarios: no regulation, 1100 µg?kg^?1 and 2000?µg?kg^?1. Because consumption distributions contained two peaks for each age category, it was assumed that two log-normal distributions for each age category were needed to achieve a better fit to the distribution models. The results of simulated DON intake using the Monte Carlo method showed that children aged 1–6 years have the highest DON intake. However, the 95th percentile of simulated intake of DON in each age group was below the provisional maximum tolerable daily intake (TDI) of 1?µg?kg^?1 body weight using any regulation scenario. The 99th percentile of simulated DON intake in the 1–6-year-old group was greater than TDI at approximately 2?µg?kg^?1 body weight. These results suggest that the current dietary intake of DON from wheat consumption does not exert a significant health effect, but we may need to reconsider the current regulation value for the 1–6-year-old age group. In addition, we may need a better method to fit the distribution to the log-normal distribution better.     

Consumer exposure to substances in plastic packaging. I. Assessment of the contribution of styrene from yogurt pots

A generic methodology for the assessment of consumer exposure to substances migrating from packaging materials into foodstuffs during storage is presented. Consumer exposure at the level of individual households is derived from the probabilistic modeling of the contamination of all packed food product units (e.g. yogurt pot, milk bottle, etc.) consumed by a given household over 1 year. Exposure of a given population is estimated by gathering the exposure distributions of individual households to suitable weights (conveniently, household sizes). Calculations are made by combining (i) an efficient resolution of migration models and (ii) a methodology utilizing different sources of uncertainty and variability. The full procedure was applied to the assessment of consumer exposure to styrene from yogurt pots based on yearly purchase data of more than 5400 households in France (about 2 million yogurt pots) and an initial concentration c₀ of styrene in yogurt pot walls, which is assumed to be normally distributed with an average value of 500 mg kg^-1 and a standard deviation of 150 mg kg^-1. Results are discussed regarding both sensitivity of the migration model to boundary conditions and household practices. By assuming a partition coefficient of 1 and a Biot number of 100, the estimated median household exposure to styrene ranged between 1 and 35 µg day^-1 person^-1 (5th and 95th percentiles) with a likely value of 12 µg day^-1 person^-1 (50th percentile). It was found that exposure does not vary independently with the average consumption rate and contact times. Thus, falsely assuming a uniform contact time equal to the sell-by-date for all yogurts overestimates significantly the daily exposure (5th and 95th percentiles of 2 and 110 µg day^-1 person^-1, respectively) since high consumers showed quicker turnover of stock.