Sir

Millstone et al. in their ill-informed Commentary “Beyond ‘substantial equivalence’”1 advocate the use of extensive animal toxicology tests on whole GM foods. But this type of testing has never been used to evaluate the safety of traditional, non-GM foods even though they often contain hundreds of unique proteins and other constituents. Such testing of whole foods would be tremendously unfocused, wasteful of laboratory animals, and unlikely to detect harmful substances, even if they were present.

The novel proteins in glyphosate- resistant soya beans are present at very low levels, for example, and their effects, if any, would not be detected by feeding whole soya beans to lab animals. A focused approach aimed at these novel proteins identified through substantial equivalence would be much more reasonable. This is the approach that has been used and it clearly demonstrates the safety of these novel components. The rest of the soya bean is identical (or substantially equivalent) to traditional soya beans and, in our view, is safe for consumers.

Millstone et al. also suggest that glyphosate-resistant soya beans exposed to glyphosate may have different composition from traditional soya beans. In our view, such concerns are not particularly valid. The composition of specific components in specific crops, GM or non-GM, varies within a range as a result of variety, environmental/climatic conditions and agricultural practices. Soya beans exposed to glyphosate are not altered in composition beyond the normal range that exists naturally. And, even if they were, animal toxicology testing would be unlikely to reveal such differences.