The World Health Organization is sceptical about the usefulness of animal feeding studies in the safety assessment of GM plants and foods. This scepticism arose in part from experience gained in the testing of irradiated food in the early 1990s. Animal studies are undoubtedly useful in the safety assessment of individual compounds such as pesticides, pharmaceuticals, industrial chemicals and food additives. It is relatively simple to feed such compounds to animals at doses sometimes far higher than humans would be exposed to and to identify any potential adverse effects on health. Foods, on the other hand, are complex mixtures of compounds. Animals can not be persuaded to eat orders of magnitude more of them than humans can, and feeding only one type of food to an animal usually reduces the nutritional value of the diet, causing adverse effects that are not related directly to the material itself. For these reasons, relating any effects on the welfare of an animal to a particular genetic modification can be extremely difficult. There is also an ethical question of whether it is right to undertake studies on animals if the results are unlikely to be meaningful.

Nevertheless, the WHO recommends animal studies if other information available on the GM plant or food is inadequate, particularly if the novel protein in the GM food has not been present in the food chain before. It also supports the use of animal testing of proteins produced by novel genes before the gene is used in biotechnology. In other words for the gene to be engineered into a micro-organism so that large amounts of its protein product can be made and purified. The individual protein can then be used in toxicology studies. Here is a list of some animal feeding studies that have been undertaken:

• Safety assessment of 1-aminocyclopropane-1-carboxylic acid deaminase protein expressed in delayed ripening tomatoes, by Reed and co-workers, published in 1996 in the Journal of Agricultural and Food Chemistry (volume 44, pages 388-494). The study analyzed GM tomato plants with delayed fruit ripening. The novel protein in the plants, an enzyme called 1-aminocyclopropane-1-carboxylic acid deaminase (ACCd), was tested under simulated mammalian digestive conditions and administered to mice at a dosage of up to 602 mg/kg of body weight (a 5000-fold safety factor relative to the average daily consumption of tomatoes). The mice showed no adverse effects.

• Evaluation of transgenic event 176 "Bt" corn in broiler chickens, by Brake and Vlachos, published in 1998 in Poultry Science (volume 77, pages 648-653). This study compared broiler chickens fed with GM insect-resistant maize with chickens fed on non-GM maize. No differences were observed between the birds.

• Feeding value of corn silage estimated with sheep and dairy cows is not altered by genetic incorporation of Bt176 resistance to Ostrinia nubilalis, by Barriere and co-workers, published in 2001 in the Journal of Dairy Science (volume 84, pages 1863-1871. This study evaluated insect-resistant GM maize and conventional maize in three separate feeding trials involving sheep and cattle. No differences were observed between the animals fed the GM diet and those fed the non-GM diet.

• Genetically modified feeds in animal nutrition 1st communication: Bacillus thuringiensis (Bt) corn in poultry, pig and ruminant nutrition, by Aulrich and co-workers, published in 2001 in Archives of Animal Nutrition (volume 54, pages 183-195). The study again involved comparisons between an insect-resistant GM maize variety and its non-GM equivalent, this time fed to poultry, pigs and cattle. Again, no differences were found between animals on the GM diet and those on the non-GM diet.

• Genetically modified feeds in animal nutrition 2nd communication: Glufosinate tolerant sugar beets (roots and silage) and maize grains for ruminants and pigs, by Bohme and co-workers, published in 2001in the Archives of Animal Nutrition (volume 54, pages 197-207). In this study pigs were fed GM gluphosinate-tolerant sugar beet and maize and compared with pigs fed an equivalent non-GM diet. No differences were found.

• The fate of forage plant DNA in farm animals: a collaborative case-study investigating cattle and chicken fed recombinant plant material, by Einspanier and co-workers, published in 2001 in European Food Research and Technology (volume 212, pages 129-134). This study investigated the fate of ingested GM plant DNA in cattle and chickens being fed a diet containing conventional maize or GM insect-resistant maize. Gene fragments from the novel gene in the GM maize were not detected in any sample from either cattle or poultry.

• Nutritional assessment of feeds from genetically modified organisms, by Flackowsky and co-workers, published in 2001 in the Journal of Animal and Feed Sciences (volume 10, pages 181-194). Digestion and feeding experiments were carried out with broilers (insect-resistant (Bt) maize), layers (insect-resistant maize, gluphosinate-tolerant maize), pigs (insect-resistant maize, gluphosinate-tolerant maize, gluphosinate-tolerant sugar beet and glyphosate-tolerant soybeans), sheep (insect-resistant maize silage, gluphosinate-tolerant maize silage) and cattle (insect-resistant maize silage). No differences were observed between GM-fed animals and animals fed on non-GM diets.

• Safety assessment of the neomycin phosphotransferase-II (NPTII) protein, by Fuchs and co-workers, published in 1993 in Bio-Technology (volume 11, pages 1543-1547). The NPTII protein, which was present in Flavr Savr tomatoes and other GM plants, including potatoes, was shown to degrade rapidly under simulated mammalian digestive conditions. It caused no adverse effects when administered to mice at a dosage of 5000 mg/kg of body weight (a million-fold safety factor relative to the average daily consumption of potato or tomato).

• The feeding value of soybeans fed to rats, chickens, catfish and dairy cattle is not altered by genetic incorporation of glyphosate tolerance, by Hammond and co-workers, published in 1996 in the Journal of Nutrition (volume 126, pages 717-727). In this study, animal feeding experiments were conducted on rats, broiler chickens, catfish and dairy cows to compare glyphosate-tolerant soybean with its conventional equivalent. The feeding value of two glyphosate-tolerant varieties was found to be comparable to that of non-GM soybeans.

• Expression of the insecticidal bean alpha-amylase inhibitor transgene has minimal detrimental effect on the nutritional value of peas fed to rats at 30% of the diet, by Pusztai and co-workers, published in 1999 in the Journal of Nutrition (volume 129, pages 1597-1603). Rats were fed GM peas and non-GM peas; the weight gain and tissue weights of rats fed either of the two diets were not significantly different from each other. (See Chapter 5 for a discussion of another of Pusztai's experiments with GM food).

• Glyphosate-tolerant corn: The composition and feeding value of grain from glyphosate-tolerant corn is equivalent to that of conventional corn, by Sidhu and co-workers, published in 2000 in the Journal of Agricultural and Food Chemistry (volume 48, pages 2305-2312). Glyphosate-tolerant GM maize was evaluated in a poultry feeding study. Results from the poultry feeding study showed that there were no differences in growth between chickens fed with the GM grain and those fed with a non-GM equivalent.

None of these studies was long-term (no food, GM or otherwise, has been subjected to a long-term study of its effects on health). I emphasise again that the World Health Organization and other expert bodies believe that their assessments based on the demonstration of substantial equivalence between GM plants and their traditional counterparts and investigations into any differences gives a much better indication of risk or lack of it.

Adapted from 'Genetically Modified Crops' by Nigel G. Halford (World Scientific Publishing Co; cccheong@wspc.com.sg)