A low-fact diet can be dangerous. Have you heard that chips, fries, and even bread can give you cancer? This "fact" has been widely reported, even though it's essentially untrue.
In April, Swedish scientists announced that a number of carbohydrate-rich foods including potato chips, french fries, and even bread contained unexpectedly high levels of a chemical known to be carcinogenic in rodents. Although the findings were, according to the Swedish National Food Administration (NFA), preliminary, the levels were reported by some international media as being "alarmingly high" and "shocking."
In spite of such inflammatory rhetoric, no governmental agency has advised the public to change its eating habits to avoid this compound. Indeed, an NFA representative was quoted as saying, "Do not stop eating these foods, but beware of what you eat..." How does it happen that consumers are first scared and then soothed about the latest health threat du jour?
The story behind this scare about acrylamide that's the suspect chemical illuminates the way scientific information (or misinformation) reaches the public, and perhaps more importantly, how scientific innuendo can affect the public's perception of food safety and risk.
Acrylamide is a small molecule that has been widely used in various industries since the 1950s. When polymerized, it is used to remove suspended solids from waste water before recycling, in oil-recovery processes, in textiles as water repellents, for various purposes in cosmetics and for technical processes like some DNA and protein analyses in biochemistry laboratories.
Is it a dangerous chemical? Yes, like most chemicals if you get enough of it. Acrylamide can get into the body by absorption through the skin, by inhalation, and in food and water. Exposure to high levels of acrylamide either acute or chronic can damage the nervous system of both animals and humans.
As far as carcinogenicity, we do know that high doses of the compound significantly increase the risk of a variety of cancers in rodents for this reason the United States Environmental Protection Agency (EPA) has labeled acrylamide a "probable human carcinogen." But there is no scientific evidence that acrylamide is a human carcinogen at any level of exposure.
People who work in industries that utilize acrylamide are naturally the ones liable to have the greatest exposure. In highly exposed workers, scientists have indeed found metabolites of acrylamide in their blood (it can bond with the hemoglobin in red blood cells), and neurological symptoms like numbness of the extremities. Thus far, though, studies to examine its human carcinogenic potential have been inconclusive.
Given that most people do not contact acrylamide in the course of their daily lives, why look for it in foods? In a study examining acrylamide levels in exposed workers, Swedish scientists found that blood levels of acrylamide in theoretically unexposed persons (the control group) were higher than expected, and they hypothesized that it might be formed when foods are cooked. Animal studies supported this theory. When rats were fed a standard, carbohydrate-rich laboratory ration, their blood levels of acrylamide were normal. But acrylamide levels rose sharply in the blood of animals fed the same food after it had been fried.
The scientists then examined a number of fried and baked human foods as a first step in evaluating their potential to raise human acrylamide levels. The results of this study were subsequently announced at a news conference by the Norway's National Food Authority, and worldwide attention was riveted on acrylamide. The World Health Organization has announced it will hold a conference on acrylamide, and the European Union is studying the results of the Swedish study. The Washington Post reported recently that NFA as well as British scientists generally confirmed the findings of acrylamide in foods cooked at high temperature.
So, with all this information and confirmation, don't we have enough to go on to be concerned and alter our eating habits? Not really. The problem is that the information about acrylamide was released to the public in preliminary form. Usually, scientists complete a study and submit it to a professional journal. Before journal editors accept a study for publication, they have the work reviewed by other experts in the field who will critique the study design, methodology, and data interpretation. Only after the author(s) satisfy the reviewers and editors will the study be published.
The acrylamide data underwent no such peer review process before public release. Further, before the scientific community accepts findings as fact, it is usual for other laboratories to repeat the same or a similar experiment, with the same results. Even a well-designed and executed study may have results that other labs cannot reproduce. A few years ago, for example, researchers from Tulane University published a paper in the prestigious journal Science that showed that combinations of various pesticides had more than additive effects on yeast cells. The study created a furor, with some environmental groups claiming that it "proved" that combinations of pesticides in the environment would have untoward effects on human health. Unfortunately for this theory, other laboratories (and even the one where the original work had been done) were unable to reproduce these results. Two years after the original paper had been published, the authors withdrew it.
While such a scenario is unusual, it illustrates the caution that should be applied to any consideration of new scientific data. But the most significant problem with the acrylamide scenario is the implicit extrapolation of animal data to humans. A basic principle of toxicology is that "the dose makes the poison." That means that almost anything that is harmless in small amounts may be dangerous at high levels of consumption. But because rodents (the most widely used experimental animals) have such short lives compared to humans, they are usually fed test compounds at levels that no human would consume. Thus, using data from animals to predict risk in humans is a tricky process, and experts may well disagree on the interpretation of results.
If a substance causes cancer or other ill effects in several species, if it has a greater effect when larger doses are given, such information helps label it a potential human health hazard. Again, replication of results is crucial. When such basic principles are ignored, some strange regulatory actions may result. For example, in 1977, the United States was going to ban the artificial sweetener saccharin, because it supposedly caused cancer in rats. It did, but only in male rats, and only because they had a rather unique way of metabolizing it a biochemical pathway that humans don't share. In actuality, saccharin escaped being banned because of a popular outcry from diabetics and others who depended on it to lower their sugar intake. Twenty-four years later, recognizing that there had been no increase in cancer even among frequent saccharin users, the government took it off its list of carcinogens, and it no longer requires a warning on its label.
Further, we should recognize that there are many chemicals, found naturally in our foods or produced by cooking, that at very high doses can cause cancer in rodents. Carrots, rich in beta-carotene, also contain the rodent carcinogen caffeic acid, as do apples, celery, and potatoes. Besides acrylamide, bread also contains the rodent carcinogen furfural. But based on body weight and dose, a 70 kg person would have to eat over 80,000 slices of bread per day for life to get the same dose that causes cancer in rats.
The moral of the story? Yes, acrylamide is a rodent carcinogen that is found in foods cooked at high temperatures. But before we start boiling our bread and pizza crust, let's take a deep breath and recognize that the data are still preliminary, and that no one has yet shown that the levels of acrylamide in foods pose any risk to humans.
Ruth Kava, Ph.D., R.D. is ACSH's nutrition expert. This article originally appeared in the National Post.
June 26, 2003
You should send this article to the editors of Redbook and tell them not to unnecessarily worry people. I am pregnant and a well-meaning co-worker showed me a brief, scary article found on page 42 of the July issue of Redbook saying that pregnant women shouldn't eat French fries or potato chips because it may make their fetuses more likely to develop cancer later in life.