SEATTLE, Wash. — Backers of Initiative 522, which would make Washington the first state to require labeling foods made from genetically modified organisms (GMOs), are careful to couch their arguments in terms of the public’s right to know, rather than public health. But the text of the measure — and the subtext of the debate — both raise the specter that eating genetically engineered food could be bad for you.
It’s a concern many people share. But sorting through the rhetoric and cherry-picked data is like wandering through the fog without a compass.
“People are positively swimming in information about GM (genetically modified) technologies,” the journal Nature noted in a special issue on the topic this spring. “Much of it is wrong — on both sides of the debate. ... (With GM crops, a good gauge of a statement’s fallacy is the conviction with which it is delivered.)”
Definitive answers are always hard to come by in science, and those who demand unequivocal proof of safety will never be satisfied. But two decades of research and analysis have converged on a conclusion that even most opponents of the technology accept: There’s no evidence that genetically engineered foods on the market today harm the health of people who eat them.
That doesn’t rule out the possibility of long-term effects, nor is it a blanket assurance that the approach is risk-free. Even ardent supporters acknowledge genetic engineering has the potential to produce crops laced with toxins or compounds that trigger allergic reactions. But so does conventional breeding.
In fact, traditional techniques have yielded several foods that caused health problems for consumers and farm workers, while there have been no such cases documented from genetic engineering.
Health concerns are only one factor people weigh in staking out their positions on transgenic foods. Environmental and pesticide effects matter to many, as do personal values and philosophies about corporate control and transparency in the food supply.
But when it comes to food safety, pro- and anti-GE camps diverge most profoundly on two fundamental questions:
• Is genetic engineering inherently more dangerous than the methods humans have been using for more than 10,000 years to mold plants and animals into safe, nutritious sources of food?
• Are there sufficient safeguards to ensure that the technology won’t someday serve up a truly dangerous entree?
Perhaps the most compelling evidence for the safety of GE crops is that millions of people have been eating them since the mid-1990s without obvious harm.
That’s not surprising, said Greg Jaffe, director of biotechnology for the Center for Science in the Public Interest. Like most food, foreign genes inserted in a plant’s DNA and the proteins they produce are largely broken down in the gut and digested.
Also, the vast majority of GE foods on the market today contain only one or two added genes, Jaffe pointed out.
A single bacterial gene is inserted in corn, soybeans, cotton and sugar beets to make the plants resistant to the herbicide glyphosate, or Roundup. The gene produces a protein that protects the crop from the herbicide’s effects, allowing farmers to spray their fields and kill only the weeds.
The other major class of GE crops are those modified with a different bacterial gene to produce a natural pesticide called Bt (for Bacillus thuringiensis), toxic only to certain insects like corn borers. Most U.S. cotton and corn has been engineered to produce Bt.
Eating corn with a built-in bug-killer sounds scary, but Bt is considered so environmentally benign — and harmless to humans — that it is widely used in organic farming. Most plants produce natural pesticides to discourage insects from eating them, and several crops have been conventionally bred to contain high levels of those chemicals.
About 90 percent of corn, soy beans, cotton and sugar beets grown in the U.S. are genetically engineered, and at least one of those crops shows up in the vast majority of chips, cereals, soft drinks, crackers and other processed foods.
Yet Americans aren’t actually eating as much GE ingredients as those statistics might suggest, Jaffe said. That’s because the GE genes and proteins are largely stripped out — along with all other DNA and proteins — during processing.
“Corn oil from GE corn is no different from corn oil from non-GE corn,” he said.
The bulk of America’s GE corn and soy actually goes to ethanol production or animal feed.
So the fact that noticeable health problems haven’t shown up in cows, pigs and chickens is another indication the crops aren’t causing harm, said Charles Benbrook, of the Center for Sustaining Agriculture and Natural Resources at Washington State University.
“If the ‘sky is falling’ scenarios were really true, it just strikes me as inevitable that we would be seeing more problems in livestock and pets,” said Benbrook, who is nonetheless critical of the safety testing and environmental impacts of GE crops.
The U.S. National Academy of Sciences, the European Commission, the American Medical Association and the scientific academies of Britain, France and Germany reviewed the evidence and concur that existing GE foods are as safe and nutritious as conventional varieties.
That consensus is one of the main reasons Jaffe’s organization, which advocates for tighter regulation of GE foods, opposes mandatory labeling.
But every GE food is different, he cautioned, so all new products need to be evaluated carefully before going on the market.
There are many ways genetic engineering can produce food that is unsafe to eat. But the human diet has always been fraught with the same kind of risks, said Toby Bradshaw, a plant geneticist and chairman of the Biology Department at the University of Washington.
Our early ancestors relied on random mutations to create grasses that held onto their seeds long enough to harvest, or grains that weren’t encased in tooth-cracking shells. Later, breeders crossed plants by hand and sorted through thousands of offspring for promising varieties. Modern molecular biology allows scientists to tag the genes they want to manipulate and create the botanical equivalent of test-tube babies, while still remaining in the realm of conventional breeding.
As a result, few crops that people eat today bear any resemblance to their wild progenitors.
“They’re no more natural than the Grand Coulee Dam or a nuclear power plant,” said Bradshaw, who uses conventional and GE techniques to breed monkeyflowers for research on plant genetics and evolution. Bradshaw was the target of activists who firebombed the UW’s Center for Urban Horticulture in 2001 in the mistaken belief that he was genetically engineering poplar trees.
In his greenhouse on campus, Bradshaw displayed two batches of monkeyflowers, which are native to mountain meadows across the West, to explain why he doesn’t consider genetic engineering inherently more dangerous than many other breeding techniques.
One batch was grown from seeds soaked in a mutagenic chemical to scramble the DNA and produce plants with a wide variety of traits. Breeders also blast seeds with gamma rays to achieve the same effect. One of the most popular varieties of red grapefruit and more than 2,000 other types of vegetables, fruits and grains — including many that are grown organically — were created through radiation mutagenesis.
Compared to genetic engineering, which inserts one or two genes, so-called mutation breeding is like taking a sledgehammer to a plant’s DNA. “It’s by far the riskiest breeding technology we use,” Bradshaw said. But it’s completely unregulated.
His second batch of monkeyflowers was engineered through the use of a bacterium that infects plants by injecting its own DNA into the plant’s chromosomes. Bradshaw replaced the bacterial DNA with the gene that gives snapdragons their scent. The technique is the mainstay of most genetic engineering, though researchers can also use a “gene gun” to propel bits of foreign DNA into plant cells.
Crosses between different species are a mainstay of conventional breeding. Pluots are the marriage of plums and apricots. Triticale, featured in whole-grain breads, results from crossing wheat with its distant cousin, rye. Rutabagas sprang from the forced mating of turnips and cabbage.
“There’s some mystical notion that the species barrier is something that shouldn’t be violated,” Bradshaw said. “But we violate it all the time.”
What’s unique about genetic engineering is how wide the gap can be between the species being crossed.
“That is different, and it’s a meaningful difference,” he said. “That’s what makes the technology both powerful and potentially risky — because you can put anything in there you want.”
A 2004 National Academy of Sciences report concluded that genetic engineering between distantly related species is more likely to cause unintended consequences than any conventional breeding method, except the sledgehammer of chemical and radiation mutagenesis.
Bradshaw uses genetic engineering to put snapdragon genes into monkeyflowers because the two species won’t cross naturally. Viruses and papayas could hardly be further apart on the evolutionary tree, yet breeders are able to “vaccinate” the fruit trees against deadly ring spot disease by slipping a snippet of viral DNA into their cells. An antifreeze gene from Arctic flounder was introduced into tomatoes to make them more cold-tolerant, though the variety was never commercialized.
In many cases, the new genes come from something humans already eat — like flounder. But in other cases, as in crosses between plants and some bacteria, the genes are introducing proteins that have not been part of the human diet before, said molecular biologist Margaret Mellon of the Union of Concerned Scientists. The UCS doesn’t oppose all genetic engineering, but has concerns about the way it’s used.
“When we start looking at different kinds of genes, we have to consider different kinds of risks,” Mellon said.
The main worry is that genetically engineered crops or animals might produce toxins or allergens, or that the insertion of the new genes could somehow disrupt the organism’s own genetic machinery, leading to changes in nutritional makeup or other consequences.
That’s where testing comes into play.
Genetically engineered foods have to jump through multiple regulatory hoops before they are allowed on American dinner plates.
The U.S. Department of Agriculture regulates field testing of GE crops. Any plant that produces a pesticide, like Bt, must go through Environmental Protection Agency review.
The Food and Drug Administration requires proof that GE crops are nutritionally equivalent to non-GE foods, but beyond that, the agency’s review is technically voluntary. In reality, every GE food has gone through the FDA process, and it would be foolish for companies to sidestep it, said Val Giddings, a former USDA regulator now with the Information Technology & Innovation Foundation.
Consumer groups and the American Medical Association want the process to be mandatory, transparent and spelled out in more detail. That’s particularly important with more complex GE crops in the pipeline, like salmon that grow faster, potatoes that don’t brown and drought-tolerant rice, Jaffe said.
“As we move to things that make up a bigger part of our diets, that’s all the more reason to have stronger oversight by FDA to be sure these risk assessments are done correctly,” he said.
In conventional breeding, genes usually move in blocks and remain in predictable locations on chromosomes. But with genetic engineering, there’s no way to know where the new genes will wind up, Bradshaw explained. They could land in the middle of an important gene and interfere with its function, or they could cause other genes to switch on or off.
If the result is a plant or animal that grows poorly or doesn’t exhibit the desired trait, it gets weeded out, just as in conventional breeding. The ones that make the cut are chemically analyzed to determine if the new gene causes natural toxins to spike, creates any dangerous new compounds or reduces nutritional value.
Problems like that have cropped up in conventional breeding, which produced a toxin-rich potato that made people sick and celery varieties so high in bug-killing compounds that farm workers who handled them broke out in rashes.
So far, genetic engineering’s record is comparatively clean. A recent review of 129 studies found no significant differences in the composition of GE crops and their non-GE counterparts.
One of the most troublesome scenarios is that genetic engineering will transfer genes from a food that causes allergies into previously harmless foods. A banana with a peanut gene hiding inside could be fatal to some people.
Companies try to prevent that by avoiding known allergens. The closest call so far came in the mid-1990s, when a company tried to boost the protein content of soybeans by adding a gene from Brazil nuts. The product was never released, because laboratory tests showed that it could trigger reactions in people with an allergy to the nuts.
GE backers cite the case as proof that the current screening system works. But critics see dangerous gaps in the safety net.
Currently, companies compare new proteins in GE foods against a catalog of known allergens. They also soak proteins in acid to see if they break down, because many allergens aren’t easily digested.
Those tests have several shortcomings, said Michael Hansen, a senior scientist at Consumer’s Union, which is critical of genetic engineering and favors labeling.
Instead of testing the actual protein produced in the plant, companies use a version produced by GE bacteria. It’s cheaper and easier that way, but the proteins aren’t identical. The acid tests also aren’t a good simulation of what goes on in the stomach, Hansen said. But most important, the tests don’t do a good job of detecting new allergens — things that aren’t already in the catalog.
Hansen pointed out that the incidence of food allergies in children has risen — from 3.4 percent in 1997 to 5.1 percent in 2011, according to the CDC — in concert with the rise of GE crops. But correlation doesn’t prove causation. In a recent forum on genetically engineered foods, University of Florida plant biologist Kevin Folta pointed out that it’s possible to graph a similar coincidence between allergies and sales of organic foods.
The Achilles’ heel of GE safety studies is their brevity. Companies are only required to feed new crops to laboratory animals for 90 days. The majority of those studies find no significant health problems, but a few hint of possible organ damage and inflammation.
The most incendiary study was published last year by French biologist Gilles-Eric Seralini, an avowed opponent of genetic engineering. It was the first time rats had been fed Roundup Ready corn for essentially their entire life spans. Seralini claimed that animals in the two-year study who ate the GE feed died earlier and were more likely to develop tumors and organ damage.
But other scientists bashed the research, pointing out that Seralini used a strain of rat prone to developing tumors in old age, didn’t have enough animals in each group to draw conclusions, and used questionable statistics.
“It’s a really bad study,” said Marion Nestle, a food safety advocate and professor of nutrition at New York University. Even though she’s a strong proponent of GE labeling, Nestle said she didn’t believe Seralini’s conclusions.
A follow-up is in the works. Whatever the outcome, the need for more long-term studies is clear, said Hansen, of the Consumer’s Union. People may have been eating GE foods for 20 years without any obvious health effects, but some diseases take decades to develop.
“We’re not saying this is going to lead to acute effects and people are going to be dropping like flies,” he said. “But it could be causing other changes, and we would have no way of knowing it.”
Labeling would allow researchers to better track possible allergic reactions and perhaps detect subtle effects, Nestle said.
But even with decades’ more study, science is unlikely to carry the day.
“When you’re dealing with belief systems, science isn’t going to work,” Nestle said. “I’m not concerned so much about the safety — but even if they are safe, they are not necessarily acceptable.”
• Sandi Doughton can be reached at 206-464-2491 or firstname.lastname@example.org.