The food safety watchdog based its conclusions on more than 11,000 food samples collected from 18 EU member states, Iceland and Norway between 1999 and 2008.
Polychlorinated biphenyls (PCBs) are a class of widespread toxic organic chemicals that accumulate both in the environment and in humans. Although the production and use of PCBs halted in1980s, large amounts remain in electrical equipment, plastic products and building materials, which can be released into the environment. Human exposure is typically through food.
The International Agency for Research on Cancer has classified PCBs as probably carcinogenic to humans.
Six cogeners
Due to the position of the chlorine atoms in the chemical, more than 200 PCB cogeners are possible, said EFSA. The agency examined six of these (Nos: 28, 52, 101, 138, 153, and 180) as indicators for occurrence of NDL PCBs as they represent around 50 per cent of substance found in food.
The EU Commission is currently discussing to lay down maximum levels for the sum of the six indicator NDL-PCBs in food and animal feed, added the report.
Of the 11,214 food samples collected, almost one fifth of the single congeners were below the limit of quantification (LOQ). PCB-153 and PCB-138 contributed the most to the sum of the six NDL-PCB congeners followed by PCB 180, PCB 28, PCB 101 and PCB 52 in that order, but relatively high variations were observed for each congener throughout food and feed groups. The contribution of PCB-153 and PCB-138 together consistently comprised at least 50 per cent of the overall sum of the six congeners in each food group.
“The highest mean of food contamination levels were observed in several fish and fish product categories followed by products of terrestrial animals,” said EFSA. “The lowest values were found in fruits and vegetables.”
Random testing
The agency flagged up weaknesses in the data collection and treatment methodologies. It said the assessment included results from both random and targeted monitoring but a clear separation of the two sampling groups had not been possible. This lack of sampling information and the irregular coverage of food groups over time meant that an accurate time trend analysis had not been possible.
“To improve the validity of any assessment of the presence of dioxins and PCBs in food and feed in Europe it is important to carry out random testing and separate reporting of a sufficient number of samples in each food and feed group,” said the report authors. “Targeted sampling during contamination incidences should be clearly indicated as such in the reporting.”