WASHINGTON, DC—Regardless of whether genetically modified foods were created by manipulating DNA in a laboratory or by traditional breeding, safety assessments should focus on the changes in a particular crop, not on the method used to create them, according to a National Research Council and Institute of Medicine report issued yesterday (July 28).

The report suggests widespread "premarket" safety assessments of any new genetically modified food based on presence of abnormal levels of compounds. It also calls for epidemiological surveillance to track any unexplained and unexpected clusters of adverse health effects that might arise only after a large population has been exposed to a particular food. But such assessments and surveillance would occur for all kinds of foods, not just GM foods.

Industry leaders met the report with praise. "Critics of the technology always want to point to the method. Just because you use the method, you've got to go through all of these hoops of regulatory policy," Michael Philips, vice president of agricultural science and regulatory policy at the Biotechnology Industry Organization (BIO), told The Scientist. "And that's not the founding principle on which we base regulations. We base regulation on the end product and how it is potentially changed though genetic modification."

But Jane Rissler, senior staff scientist at the Union of Concerned Scientists' Food and Environment Program, worried that the report's recommendations will dilute assessment efforts. "This committee is going back and obscuring the issue by trying to bring in all of these traditional techniques," she told The Scientist.

Genetically engineered foods should be subject to more scrutiny because they represent the greatest uncertainty, she said. "If we had plenty of resources to look at food safety, I would say focus on everything," said Rissler. "But given that we don't, we have to focus where the greatest concern is."

Henry Daniell, a professor of molecular biology at the University of Central Florida, said that in principle, the "process" could be very important. If, for instance, a gene for insect resistance were inserted into a plant's chloroplast genome rather than, as is typical, its nucleus, that gene would not travel in the plant's pollen and have a chance of affecting the environment by outcrossing with related crops or weeds.

Daniell said he also sees important scientific distinctions between conventionally bred foods and genetically modified foods. For example, scientists routinely use antibiotic resistance genes as markers to determine whether their gene of interest has been inserted. Theoretically, products of these tag-along genes, not present in conventional breeding, could be allergens. However, said Daniell, as yet there's no conclusive proof that transgenic foods harbor such allergens.

Panel chairman Bettie Sue Masters, a professor of chemistry at the University of Texas Health Science Center in San Antonio told The Scientist that she envisions a comprehensive "feedback loop" in which industry and regulators assess food safety for consumers while, at the same time, health effects in the marketplace are efficiently communicated back to those regulating and developing the technology.

BIO's Philips said that the biotech industry already complies with the assessments suggested by the National Academy of Science (NAS) panel in consultation with the Food and Drug Administration (FDA) (though, said Philips, a widespread post-commercialization evaluation only occurs by agency request, and hasn't yet happened). Rissler said that such consultations are voluntary and that there are no stringent FDA regulations.

"The Program currently underway is a voluntary consultation program," James Maryanski, Biotechnology Coordinator at the FDA's Center for Food Safety and Applied Nutrition, told The Scientist. "It's prudent practice for new companies to consult with us on new technology."

Food additives, but not foods, require premarket approval. However, if a company inserts a gene into a plant, and that gene expresses a protein that's not generally recognized as safe, it's then considered a new food additive, and would require FDA approval, Maryanski said. The FDA can take legal action against a company that, after skipping consultation, releases a food that causes adverse health effects.

The NAS report also calls for new research and methodologies to discover and track the potential ill effects of modified foods—for example, profiling techniques in animal models that would link altered gene expression with metabolic components, and extensive databases of the biological composition of many different species. The compounds in a novel genetically modified plant could then be compared with a large control group of its safe, "conventional" cousins.

Last month, the United Nations' Food and Agriculture Organization published guidance on how to assess the risk plants face from living modified organisms, including genetically modified seeds, cuttings and tissue cultures, as living parts of plants.

Correction (posted July 30): The sixth and seventh paragraphs of this story have been altered to clarify several points made by Henry Daniell. The Scientist regrets the errors.