Furan is an organic compound with aromatic properties which has been found to be carcinogenic in animal studies. It can form during the heat-treatment of food products, and contributes to the taste and smell of a prepared food product.
Red flags were first waved over levels in tinned and canned foods in 2004, when the US Food and Drug Administration reported detecting higher levels than previously thought. It said this was not due to an increase in furan levels, but that new analytical techniques now allow for better detection. The reason for the formation of furan is not yet certain.
This discovery led EFSA to set up a panel of science experts to gather more information on the chemical, and in 2006 EFSA held a joint workshop with DG Health and Consumers, and the European Commission Joint Research Centre. A call was issued for member states to start collecting data in 2007 and 2008.
EFSA published its first report on monitoring last year, and this week has released the follow-up which includes all data from 2004 to 2009.
Eighteen member states sent in data on 4,186 samples across 21 food categories. However only 8 per cent of the samples were on foods as consumed. Because furan is highly volatile, the food preparation process can have an impact on levels between opening of a can and serving it up on plates.
Moreover, samples appear to be concentrated on certain food categories, such as coffee and baby and infant formula – although these are deemed to be important, especially as the latter could comprise the entire diet of a child.
Other foodstuffs included were cereal products like puffed rice, fish products like mackerels and sardines in tomato sauce, meat products like canned duck with lentils or rabbit with prunes, soups like tomato soup and in gravy, and milk products.
“Future testing of furan by Member States should preferably target food products where limited results are available and comprise, if possible, the sample analysed as purchased followed by the same sample analysed as consumed indicating the exact food preparation method used,” said EFSA.
EFSA’s panel then sub-divided some of the categories. Baby foods were sub-divided into 6 groups, and coffee into five.
In coffee, mean values ranged from 6,900 µg/kg in ‘coffee roasted ground’ down to 602 µg/kg for ‘coffee instant’, 1,807 µg/kg for ‘coffee roasted ground’, 1,855 µg/kg for ‘coffee non specified’, 3,611 for ‘coffee roasted bean’, 104 µg/kg for ‘coffee ready-to-drink’.
However EFSA said: “There is obviously a dilution effect in preparing the coffee, however, little detail was provided on the brew recipes and no information was provided on the type of preparation.”
The mean furan content in infant formulae was 3.2 µg/kg, but in the different baby food categories it ranged from 5 µg/kg for baby food containing only fruits to 40 µg/kg for baby food containing either meat and vegetables or vegetables only.
Another interesting observation was made for milk products. Milk based processed food showed low mean furan content of 6 µg/kg, but a maximum furan content of 80 µg/kg was found in sweetened condensed milk.