Novel foods, whether derived by GM, or newly imported from another country or produced by a novel industrial process from a novel source like Quorn, cannot be evaluated as if they were drugs. A drug can be tested for its toxicity, first in animals and then in humans, by using very large doses, say 100 or 1,000 times higher than that expected in practice.

This is impossible with novel foods as such, because one cannot feed people the novel food at 10 times the normal amount they might eat. Not only would they balk at the prospect but such a large amount would greatly disturb their livers, not because it was intrinsically unsuitable but as a result of the sheer volume.

So other methods have to be used. Specifically, the gene product itself can be tested for toxicity at much higher concentrations than will appear in the food - say 1,000x - and this was done for the one that makes GM soybean resistant to the herbicide “Roundup”. No toxicity was detected.

Animal testing is extensively employed. Usually it is possible to use doses only a few times higher than in the normal diet. The most useful method, however, is to compare the new food with one that had been eaten for many years since the reactions of a wide variety of different peoples will be known: specific investigations are made of chemical composition, nutritional quality, toxicity and allergenicity (as well, of course, of the effect of the specific genetic intervention). If no difference from the conventional plant can be detected the new variety is described as being “substantial equivalent”. If all these tests indicate no problem, limited human trials are possible with volunteers. This was the process followed both with Quorn (made from a mould) and when flour from lupins became available for sale.

In the lupin case, allergy was the worry. Although there were no contraindications, plans for releasing the flour for use in a limited area involved alerting all the local allergy clinics and GPs. In the event there were no problems.

It is difficult if not impossible to do long-term trials in humans lasting many years. Over a protracted period, a thousand or more people (a large number because we are looking for small effects), representative of the populations as a whole, would have to be fed their normal diet plus the novel food, while another equally large population would have to eat exactly the same diet without the novel food. That simply cannot be done: people will not tolerate diets like that for months or years on end.

It has also been suggested that any adverse effect of a GM food could be picked up by monitoring a random population for illness and seeking to correlate any illness with their diet using the supermarket loyalty cards to assess the diet. This was ruled out as an invasion of privacy.

The fact that 300 million in North America have been consuming GM foods for ten years without a single substantiated case of harm, while not a rigorous scientific experiment, does indicate that the dangers are, to say the least, muted.

Source:

V. Moses and M. Brannan (2001). One hundred percent safe? GM foods in the UK. CropGen (click to download)

<<<back