Results of the monitoring of dioxin levels in food and feed

Environmentally persistent dioxins and dioxin-like compounds include 29 congeners of dioxins, furans and polychlorinated biphenyls (PCB) with similar toxic effects, their quantification commonly expressed as toxic equivalent units according to their varying potency. While the amount of those compounds in the environment has declined since the late 1970s, there is a continued concern because of their accumulation in the food chain, particularly in animal fat. In 2002 the European Commission prescribed a list of actions to further reduce the presence of dioxins and dioxin-like PCBs and later introduced action and maximum levels with random monitoring by Member States. A total of 7,270 samples collected in the period 1999-2008 from 19 Member States, Norway and Iceland were analysed in detail. Dioxin and furan congeners comprised between 30% and 74% of the total concentrations depending on food or feed group, while mono-ortho PCBs comprised between 15% and 45% of the dioxin-like PCBs. The highest mean levels of dioxins and dioxin-like PCBs in food expressed on fat basis were observed for ‘liver and products thereof from terrestrial animals’ and on whole weight basis for ‘fish liver and products thereof’. In feed the highest levels were found in ‘fish oil’. An overall 8% of the samples exceeded different maximum levels and a further 4% exceeded some action levels. However, some of these samples clearly originated from targeted sampling during specific contamination incidences and there were large variations between groups. Changing the basis for calculating toxic equivalent units to the new recommendations issued by WHO in 2005 will overall result in 14% lower values with the extent of the difference highly variable across food and feed groups. To ensure accurate assessments of the presence of dioxins and dioxin-like PCBs, continuous random testing of a sufficient number of samples in each food and feed group is recommended.

© European Food Safety Authority, 2010

This report published on 22 July 2010 replaces the earlier version published on 31 March 2010

Dioxins and dioxin-like compounds include a range of toxic and environmentally persistent substances. The terms most often refer to 29 congeners of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans and dioxin-like polychlorinated biphenyls (PCB). Dioxin and furan congeners are formed as unintentional by-products during combustion processes such as waste incineration and forest fires, as well as during some industrial processes such as paper pulp bleaching and the manufacturing of chlorinated pesticides. PCBs are synthesised by direct chlorination of biphenyl and can be divided into different groups according to their biochemical and toxicological properties. Non-ortho and mono-ortho substituted PCBs show toxicological properties that are similar to dioxins. They are therefore often termed ‘dioxin-like PCBs’.  The other PCBs don’t exhibit a dioxin-like toxicological profile and are therefore termed ‘non-dioxin-like PCBs’. PCBs have been used in a variety of applications such as dielectric fluids in transformers and as heat transfer fluids because of their non-flammability and electrical insulation properties. The production and use of PCBs have been discontinued in most countries since a ban on their manufacturing, processing and distribution was introduced in 1985, but large amounts remain in electrical equipment, plastic products and buildings.

While the amount of dioxins and dioxin-like compounds in the environment has declined since the late 1970s, there is a continued concern about the safety of the food supply and the potential adverse health effects of exposure to this group of substances. Because of their lipophilicity, together with their persistency in the environment, dioxins have accumulated in the food chain, particularly in animal fat, dairy products, and fish.

The presence of dioxins and dioxin-like PCBs is expressed as toxic equivalents (TEQ) after multiplication of congener-specific concentration levels with toxicity equivalency factors (TEF) developed based on their relative toxicity compared to 2,3,7,8-TCDD. The current European legislation is based on TEFs set by the World Health Organisation (WHO) in 1998 with the results expressed as TEQ_WHO98. New TEFs were suggested in 2005 with the results expressed as TEQ_WHO05.

In 2002 the European Commission prescribed a list of actions to further reduce the presence of dioxins and dioxin-like PCBs and later introduced regular monitoring by Member States of food and feed, including, if possible, also non-dioxin-like PCBs. Data on the presence of 17 congeners of dioxins and furans, and 12 congeners of dioxin-like PCBs in food and feed have been reported on a regular basis to the Commission. In April 2008 the Commission handed the collected information to the European Food Safety Authority (EFSA) for assessment.

A total of 7,270 samples collected in the period 1999-2008 from 19 Member States, Norway and Iceland were analysed in detail. The percentage of samples below the limit of quantification (LOQ) varied considerably at the congener level. Overall, the percentages of censoring, defined as the proportion of non-quantified (<LOQ) observations, varied sizeably depending on how results were expressed: on a fat basis (about 40%), on a whole weight basis (about 30%), or for feed on ‘12% moisture’ basis (about 60%).

The highest mean levels of dioxins and dioxin-like PCBs in food were observed for ‘Fish liver and products thereof’ (32.6 pg TEQ_WHO98/g) and ‘Muscle meat eel’ (6.7 pg TEQ_WHO98/g) expressed on whole weight basis, and for ‘Liver and products thereof from terrestrial animals’ (5.7 pg TEQ_WHO98/g) expressed on fat basis. The highest level in feed was found in ‘Fish oil’ (10.0 pg TEQ_WHO98/g) expressed on 12% moisture basis. 

Ad hoc analyses were conducted for the groups ‘Meat and meat products ruminants’, ‘Muscle meat fish and fish products excluding eel’, ‘Raw milk and dairy products including butter’, and ‘Hen eggs and egg products’ to evaluate the influence of species, geographical or production differences. No clear conclusion could be drawn for the ruminant meat group because of low sample numbers for some species. For the food group muscle meat of fish and fish products excluding eel there were differences in dioxins and dioxin-like PCBs between the different sub-groups, with mean values ranging from 1.20 (‘Farmed trout’) to 7.99 (‘Salmon’) pg TEQ_WHO98/g expressed on whole weight basis. Mean levels of dioxins and dioxin-like PCBs in herring were overall higher for samples collected in countries from the Baltic Sea area (8.64 pg TEQ_WHO98/g) compared to countries outside the Baltic area (2.30 pg TEQ_WHO98/g). In the food group milk and dairy products, slightly lower mean levels of dioxins and dioxin-like PCBs were found when moving from farm (1.27 pg TEQ_WHO98/g) and bulk milk (1.30 pg TEQ_WHO98/g) to retail milk (0.95 pg TEQ_WHO98/g) expressed on fat basis. This could be related to a slight dilution effect at retail level in the mixing of milk from different origins or to a targeting of possible suspect samples at farm level. Too few egg samples had been assigned a specific production method to serve as a basis for any detailed analysis.

The percentage of results exceeding different maximum levels for dioxins and dioxin-like PCBs set by legislation was on average 8% with a further 4% exceeding some action levels, but there were large variations between groups. Overall, a lower percentage of results exceeded maximum levels set for feed than for food. It is important to bear in mind that a varying proportion of product testing reflects targeted and not random monitoring. This has the potential of introducing a degree of uncertainty and bias in the evaluation of background levels of dioxins and dioxin-like PCBs in food and feed, as higher total values are expected in targeted compared to random samples.

The impact of changing the basis for the calculation of TEQ from TEF_WHO98 to TEF_WHO05 was evaluated with levels of dioxins and dioxin-like PCBs using the latter being overall 14% lower than levels using the former. This difference was mainly due to changes in TEFs for mono-ortho PCB and furan congeners with little change in dioxin and non-ortho PCB congeners. However, there were large variations observed for different food and feed categories and between products within food and feed categories.

Dioxin and furan congeners comprised between 30% and 74% of the total concentration of dioxins and dioxin-like PCBs depending on the food or feed group. There was considerable variation within food and feed groups. Mono-ortho PCBs comprised between 15% and 45% of the concentration of dioxin-like PCBs. This proportion was considerably lower when using the TEF_WHO05 rather than TEF_WHO98 as the basis for calculating the TEQ.

The current results clearly include results from both random and targeted monitoring although not specifically stated and should be interpreted with some caution. The lack of such sampling information and the irregular coverage of food and feed groups over time did unfortunately not allow for a time trend analysis to be performed. To improve the validity of any assessment of the presence of dioxins and dioxin-like PCBs in food and feed in Europe random testing and separate reporting of a sufficient number of samples in each food and feed group is important. Targeted sampling during contamination incidences should be clearly indicated as such in the reporting.

Polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans, dioxin-like polychlorinated biphenyls, toxicity equivalency factors, toxic equivalents, food, feed