Following a request of the European Commission and the need for a joint assessment identified by the European Environmental Agency, the EFSA estimated the chronic dietary exposure to lead in the European population, updating the results of the 2012 assessment. The highest mean lead occurrence was observed for food category 'Meat and meat products', strongly affected by extreme lead concentrations found in game mammals' meat, the impact of which was assessed via two exposure scenarios including (A) an excluding (B) them. In the scenario A, the highest mean exposure upper bound (UB) and highest 95th percentile exposure UB were estimated respectively in toddlers (1.80 μg/kg body weight (bw) per day) and in infants (3.36 μg/kg body weight (bw) per day). Across the different age classes, the main contributors to the dietary exposure to lead in adults were the meat-based foods, mostly 'Cured ripened raw sausages', other types of sausages, game mammals' meat and 'Bread and similar products'. Ad hoc exposure scenarios for high consumers of game animals' meat showed ⁓six-fold higher exposure as compared to the total population when extreme lead levels were considered. The authority also investigated the time trend in the occurrence. A linear shape was considered appropriate to describe the time trend. A random effect was used to account for similarity of measurements in the same country. The lead occurrence showed a slight decrease for most of the food categories with an increase observed for the 'Legumes, nuts, oilseeds and spices' (+0.32 μg/kg per month), 'Vegetables and vegetable products' (+0.17 μg/kg per month) and 'Coffee, cocoa, tea and infusions' (1.28 μg/kg per month). Uncertainties were identified and recommendations listed to improve future dietary exposure assessments to lead.
Recently, the European Environment Agency (EEA) calculated the burden of disease from lead using Human BioMonitoring (HBM) data and found levels in blood samples of people from various countries comparatively high (i.e. over 5 μg/dL, the clinical reference level recommended by the WHO). The need for a comprehensive understanding of the sources and routes of exposure to lead in the European population triggered a collaborative approach between EEA and EFSA that was reflected in an official request from the European Commission. Following this request, the EFSA assessed the dietary exposure to lead in the European population, considering the occurrence data in food and drinking water, submitted after the publication of the 2012 EFSA scientific report. The EFSA was also asked to provide an overview of the available occurrence data on lead in food and to investigate possible trends of occurrence over the years using all the available data.
In the current report, a total of 288,239 analytical results on lead were initially considered for the dietary exposure assessment, among them 9922 corresponding to drinking water and 278,317 to different types of food. Samples were collected across Europe between 2013 and 2023. Consumption data from 23 different European countries and a total of 55 different dietary surveys were used to estimate the chronic dietary exposure to lead.
The highest mean lead concentrations occurred for 'Meat and meat products', impacted by several very high lead concentrations found in game mammals' meat with highest concentrations of 3.7 g/kg measured in wild boar meat and cured ripened raw sausages with the highest concentration of 0.36 g/kg. High mean lead concentrations were observed also for food supplements and seasonings.
The impact of a set of extremely high lead levels observed in game mammals' meat was assessed via two exposure scenarios with (scenario A) and without (scenario B) the extreme lead levels. The highest dietary exposure was estimated in the young population (infants, toddlers and other children). In the scenario A, the highest estimated mean exposure upper bound (UB) level was in toddlers with a maximum exposure of 1.80 μg/kg bw per day and the highest 95th percentile exposure UB was observed for infants with estimate of 3.36 μg/kg body weight (bw) per day. Lower exposures were estimated when the extreme lead levels measured in game mammals' meat were excluded (scenario B). The main contributors to the dietary exposure to lead in adults were the meat-based foods, mostly 'Cured ripened raw sausages', other types of sausages, game mammals' meat and 'Bread and similar products'. In few surveys also other foods, e.g. escaroles, wine, pome fruits and salads, etc., were important contributors. Particular foodstuffs played a relevant role in the dietary exposure to lead in the young population, including 'Cereal-based food for infants and young children' in the infant population and in several surveys also herbal infusions, birds meat, pome fruits, milk, fruit juices and nectars, carrots, celery leaves and fruiting vegetables.
As compared to the 2012 EFSA scientific report, the maximum lower bound (LB) mean and 95th percentile exposure estimates to lead across dietary surveys are now lower with the maximum means and 95th percentile estimates across the different age classes around 1.3-1.7 times lower as those reported in the 2012. This difference is likely due to some methodological aspects such as improvement of the linkage between consumption and occurrence, thanks to the availability of additional information and updated food classification system. The difference is also expected to stem from changes in the food consumption data and lower occurrence levels of lead reported in last 10 years.
Different ad hoc dietary exposure scenarios were conducted to complement the general exposure scenario. Among those, a scenario calculated for high consumers of game mammals' meat showed an important ⁓'six-fold' level of exposure as compared to the total population with estimate up to 5.18 μg/kg body weight (bw) per day in adults. However, this outcome is considerably impacted by inclusion of the set of extremely high lead concentrations measured in deer and wild boar meat; after removal of the extreme lead levels, the maximum LB for high consumers of game mammals' meat was ⁓30% higher as compared to total population.
The trend analysis was performed using the data available in the EFSA occurrence database. Although years from 2003 to 2023 were available, the first and last years of the interval were excluded due to the very limited sample size (10 observations in total). Before performing the analysis, a data cleaning was carried out whose strategy consisted in identifying very extreme values and excluding them based on criteria that are applied systematically across the years. Since the data on occurrence cover a timespan of almost 20 years, it was not possible to ask clarifications to the data providers. After cleaning, the total sample size of the occurrence data was equal to 418,273 and covered the time span 2004-2022 (19 years).
The presence of a time trend in the occurrence of lead was first investigated graphically to identify the most appropriate shape. A linear trend and two types of non-linear trends (i.e. a locally estimated scatterplot smoothing - LoESS - and a natural cubic spline) were explored. Based on the preliminary analysis, a linear shape was considered appropriate to describe the time trend and a linear mixed model (Brown & Prescott, 2006) was used with time as the fixed effect and a random effect to account for the correlation among multiple observations taken in the same country. Each food category was analysed separately using the Foodex2 level 1 classification system. The monthly trend was considered in addition to the yearly trend to capture more accurately the variation over time. Two models were run, with and without the extreme values, to assess whether the cleaning had improved the fit of the model.
The goodness of fit was assessed using the conditional R that describes the proportion of total variance explained by the fixed and random effect. Despite the preprocessing of the data improved the fit, the latter remained generally poor with the conditional R for the monthly trend model ranging from 0.05 for the categories 'Sugar and similar, confectionery and water-based sweet desserts' and 'Milk and dairy products' and 0.45 for 'Water and water-based beverages'. This result was also confirmed by the analysis of residuals that highlighted several potential outliers that could not be addressed in detail for the reasons previously discussed.
The lead occurrence showed a slight decrease over time for the majority of the food categories (Tables 12 and 13 for the monthly trend, Table 14 and Figure D.3 in Annex D for the yearly trend) with the greatest drop observed in 'Products for non-standard diets, food imitates and food supplements' followed by 'Seasoning, sauces and condiments' (0.59 μg/kg and 0.28 μg/kg decrease per month for the two categories, respectively). The food items characterised by an increase in the lead occurrence are the 'Legumes, nuts, oilseeds and spices' (+0.32 μg/kg per month), 'Vegetables and vegetable products' (+0.17 μg/kg per month) and 'Coffee, cocoa, tea and infusions' (1.28 μg/kg per month). However, for those food items, the intercept is quite uncertain confirming the possibility that a non-linear model might be more appropriate in these cases.
To investigate more in detail the time trend of the lead concentrations in the three food categories showing a potential increase over time, the Foodex2 level 2 was considered (summary statistics provided in Annex E). In the category 'Legumes, nuts, oilseeds and spices', none of the related food categories showed an evident increasing trend. However, the lead concentration in 'spices' highlighted a tendency to exceed the level of 500 μg/kg and in some cases reaching 1000 μg/kg more often during the later time period than earlier despite the parameters of the distribution not reflecting an evident trend. Among the 'vegetables and vegetable products', a slight increase is observed in the 'processed or preserved vegetables and similar' with a sustained tendency of levels of 500 or above in more recent years, as partly reflected also by the mean value. In the same category, the lead concentration in 'herbs and edible flowers' showed a slight increase after 2020 and is characterised by a large spread with levels ranging from 0 to almost 1300 μg/kg. The possible increase of contamination from lead in the class of 'coffee, cocoa, tea and infusion' seems imputable to the 'ingredients for coffee, cocoa, tea and herbal infusions' that highlight a relatively stable increase over time of almost all the distributional parameters.
The main uncertainties associated with the dietary exposure estimations refer to the impact of using the substitution method to treat the left censored data, to the lack of information (consumption and occurrence) on some lead-containing ingredients in specific food groups, to the effect of food preparation on the lead levels, the lack of occurrence data in human milk and the presence of high levels of occurrence that are suspected to be outliers but could not be verified. Recommendations are provided to allow improving future dietary exposure assessments to lead. These recommendations are mostly focused on asking data providers to submit analytical data following the requirements as specified in the Chemical monitoring reporting guidance, including accurate description of the samples and information on ingredients (e.g. game mammals' meat), to use validated analytical methods with adequate sensitivity and to further investigate the reason for a decreased monitoring in drinking water, in particular in tap water. With respect to game mammals', it is recommended to accurately report the production method, whether the animal was hunted or farmed. It is recommended that human biomonitoring data are collected on human milk with the purpose of providing occurrence data for the estimate of exposure to contaminants (including lead) for breastfed infants.