Sir

One way to allay public concerns and to find out more about the effect of genetically modified organisms (GMOs) would be to investigate more fully their use in animal feeds. Much money is spent on determining the safety of GMOs as human foods, but would it not be cheaper, easier and more ethical to test animal feeds first?

Large quantities of plant materials, produced by genetic engineering, are destined as raw materials for animal feed: 85 per cent of maize, for example, is used as animal feed or as agro-industrial by-products. Most soya beans are used as protein-rich meal for animals; almost two-thirds of unginned cotton, and most rape seeds and tomato pomace are used as or in feedstuffs. These crops are among the first GMOs submitted for licensing, and will end up in the human food chain. It is obviously more convenient for research to be done on animal feeds rather than on human food.

The central concept in animal nutrition is ‘nutritive value’ which is influenced by the presence of undesirable substances, including the potential transfer of harmful factors introduced into the DNA of plants during their conversion into GMOs. Companies base their safety criteria on the principle of ‘substantial equivalence’ between the engineered and the corresponding conventional plants. To measure this, they generally use chemical, in vitro and in vivo analyses. Chemical methods compare the sequence of amino acids of the introduced protein with those of known allergenics; in vivo methods use small laboratory animals for acute oral toxicity tests of relatively short duration. Although these methods are useful tools, one cannot safely extrapolate between species. Biology is often unpredictable: for example the antibiotic cross-resistance to ampicillin in humans. In GMO plants resistant to herbicides, a complex is created between the ‘factor introduced for resistance’ and the ‘herbicide’. The possibility cannot be ruled out that this complex could be broken down during digestion in the gut or during fermentation, resulting in release of the herbicide.

In addition to the need for labelling and an increased role for legislation and monitoring (guidelines), there is a strong need for research in ‘evaluation’. Companies have to demonstrate that GMOs are both effective and non-toxic. Risk assessments are essential to ensure the latter. Study of feeds and farm-animal nutrition for at least one reproductive cycle is also needed. If the health of the animals is not harmed as a result of these tests (which should be done in government-funded institutions), the public is more likely to be reassured. Companies would be in a better position to convince the public of the safety of GMOs.