This fascinating material is excerpted from The Big Fat Surprise: Why Butter, Meat & Cheese Belong in a Healthy Diet by Nina Teicholz, reprinted with permission of Simon & Schuster. The effort to remove trans fats from the diet has been largely successful; unfortunately, what restaurants, fast food establishments and food processing companies are now using may be much worse. For a review of the book, click here.
In late 2012, as I was researching the latest news on trans fat replacements, Gerald McNeill, vice president of Loders Croklaan, which is one of the country’s largest suppliers of edible oil, told me something scary. He explained that fastfood chains including McDonald’s, Burger King, and Wendy’s have swapped out hydrogenated oils and started using regular vegetable oil instead. “As those oils are heated, you’re creating toxic oxidative breakdown products,” he said. “One of those products is a compound called an aldehyde, which interferes with DNA. Another is formaldehyde, which is extremely toxic.”
Aldehydes? Formaldehyde? Isn’t that the stuff that’s used to preserve dead bodies?
He went on to tell me how these heated, oxidized oils form polymers that create “a thick gunk” on the bottom of the fryer and clog up the drains. “It’s sticky, horrible! Like a witches’ brew!” he exclaimed. Partially hydrogenated oils, by contrast, were long-lasting and stable in fryers, which is of course why they were favored. And beef tallow, McDonald’s original frying fat, was even more stable.
McNeill’s company was a subsidiary of a giant Malaysian corporation that sold palm oil, so I wondered at first if he wasn’t just vilifying the competition. Then I called Robert Ryther, a senior scientist at Ecolab, the giant industrial cleaning company that services nearly all the major national fast-food restaurants, and he confirmed the “gunk” issue. “It builds up on everything. It’s like paint shellac . . . anywhere from a real hard, clear coating to a thick, gooey material, like a white silicone lubricant that you use on car engines, with a Crisco-type feel to it.” The gunk, he said, is the result of a hot oil mist coming off the fryer and then collecting on cold surfaces all over the restaurant—in mixers, ovens, and vents and on the floors and walls. Within a day, it would start building up. “Literally,” says Ryther, “we’d go into [restaurants], and people would say that we’ve been trying to get rid of this stuff for three weeks using sand blasters or hand scraping.”
Ryther told me that these unstable products from oils would also accumulate on the uniforms of fast-food workers, which, when heated in clothes dryers, had been known to spontaneously combust. And fires would start in the back of the trucks carrying the uniforms to be cleaned. Even after the laundry was clean and folded, it would sometimes catch fire, Ryther told me, “because the oxidation products are continuing to react in very small amounts. You’re never going to get it all out, and they will generate heat.” Ryther started seeing this problem in 2007, shortly after restaurants went trans-free and converted their frying operations over to regular vegetable oils.
Ryther developed a product called Exelerate ZTF, which converts the shellac-like substance back into oil so that it can be cleaned off. The process is more expensive than previous solutions, however, and also uses stronger chemicals, so it’s not a job for untrained employees. And pretty much all restaurants, large and small, are dealing with this, says Ryther. “McDonald’s had this problem. Anybody that has a fryer has this problem.”*
An obvious health question is whether these substances might also damage the lungs of patrons and restaurant workers.† And in fact, rates of cancers of the respiratory tract have been found to be higher among chefs and restaurant workers in Britain and Switzerland, where the subject has been studied.‡ However, these studies did not track the type of cooking fat used and were confounded by the fact that the stoves themselves also emit damaging microparticles. Nevertheless, the highest-level report on cancer and heated oils to date, published in 2010 by the International Agency for Research on Cancer (IARC), which is part of the World Health Organization, determined that emissions from frying oils at the temperatures typically used in restaurants are “probably” carcinogenic to humans.
The problem, as we know, is that these regular vegetable oils oxidize easily, and heat speeds up the reaction, especially when heated over periods of hours, as typically occurs when these oils are used in restaurant fryers.
The linoleic fatty acid in these oils starts a snowballing chain of reactions. Linoleic fatty acid comprises 30 percent of peanut oil, 52 percent of soybean oil, and 60 percent of corn oil, and it degrades into oxidation products such as free radicals, degraded triglycerides, and others; in one analysis, a total of 130 volatile compounds were isolated from a piece of fried chicken alone.* And while the IARC report looked only at the effects of particles that were airborne, it said nothing about those absorbed into foods fried in these oils. And it seems likely that the impact of these oxidation products is far greater when they are eaten—and digested.
Oil chemists began discovering these compounds in the mid-1940s, when vegetable oils first came to be widely used, and published a large body of work showing that heated linseed, corn, and especially soybean oil were toxic to rats, causing them to grow poorly, suffer diarrhea, have enlarged livers, gastric ulcers, and heart damage, and die prematurely. In one experiment, a “varnish-like” substance was found in the rat feces—which caused the animals themselves to be “stuck to the wire floor” of the cages. The oil in some of these experiments was heated to temperatures higher than those typically used in restaurant fryers, but the “varnish” was likely to have been an oxidation product in the same family as those shellac-like substances turning up in fast-food restaurants of late.
One would think that these disturbing early findings would have generated a great deal more research and discussion, especially since the AHA started recommending these polyunsaturated oils to the public in 1961. However, one of the few U.S. researchers warning authorities not to jump into embracing the oils so quickly was the chemist Denham Harman, a founder of the hypothesis that free radicals cause aging. The scientific literature on the negative effects of these oxidation products was convincing enough, wrote Harman in a letter to The Lancet in 1957, that “the present enthusiasm” for these unsaturated oils should “be curbed” pending additional study of the possible adverse health effects of this dietary change.
Yet since then, publications and international meetings on the topic have been rare, even as research continued to turn up worrisome results. At a symposium on the topic attended by industry scientists in 1972, for instance, teams of food chemists from Japan reported that heated soybean oil produced compounds that were “highly toxic” to mice. A pathologist from Columbia University also reported that rats fed “mildly oxidized” oils suffered liver damage and heart lesions, compared to rats fed tallow, lard, dairy fats, and chicken fat, which showed no such damage. Most of this research was published in obscure, highly technical journals that nutrition experts rarely read, however; and in the U.S., diet-and-disease researchers were instead focused almost exclusively on cholesterol anyway.
Interest in these oxidation products picked up in the 1990s, when an especially toxic one, called 4-hydroxynonenal (HNE), was identified by a group of researchers at the University of Siena, Italy. This was one of those aldehydes that Gerald McNeill had mentioned to me. Hermann Esterbauer, an Austrian biochemist, is credited with discovering the general category of aldehydes as peroxidation products in 1964, and in 1991 he took stock of the field. His review is considered a landmark, and it is, frankly, a little terrifying to read. Esterbauer goes through the evidence that aldehydes are very chemically reactive, causing “rapid cell death,” interfering with DNA and RNA, and disturbing basic cell functioning. He meticulously lists all the research to date showing that aldehydes cause extreme oxidative stress to every possible kind of tissue, with a “great diversity of deleterious effects” to health, all of which were “rather likely” to occur at levels normally consumed by humans.
Aldehydes are “very reactive compounds,” says the Hungarian-born biochemist A. Saari Csallany, who studied with Esterbauer and is the main researcher of these compounds in the United States. “They are reacting constantly. From one minute to the next, they have decomposed and changed into something else.” In fact, one of the reasons that aldehydes were not more studied until relatively recently is that they were hard to measure accurately, and researchers therefore did not know that they occurred in such large amounts. Csallany refined the ability to detect HNEs and showed that they were produced by a range of vegetable oils, at temperatures well below those regularly used for frying and long before the oils start to smoke or smell, which are the alarm bells normally employed to signal that the oils are going bad.* Many oxidation products, including HNEs, are not detected by the standard tests restaurants use to monitor their oils.
One of Csallany’s recent projects involved buying fries at six fast-food restaurants in Minneapolis near her office at the University of Minnesota, which led to the discovery that people could easily eat “quite a lot” of these toxic compounds (13.52 μg HNE per 100 grams of fries). She would like to do more studies, but she says the NIH and USDA have shown minimal interest in funding this topic.
The proliferation of research has mostly been in Europe over the past decade. The strongest evidence now points to HNE’s role in atherosclerosis, says Giuseppi Poli, a biochemist at the University of Turin who co-founded the International 4-HNE Club in 2002, which now meets every two years. HNEs cause LDL-cholesterol to oxidize, which is thought to be what makes that kind of cholesterol dangerous. And the evidence implicating HNEs in the development of neurodegenerative diseases like Alzheimer’s is also strong, he says. Moreover, HNEs so reliably create oxidative stress in the body that they are used as a formal marker for the process.
This kind of stress was observed in an experiment on mice fed a type of aldehyde called acrolein, named for its acrid smell when produced by overheated oils. It is also present in cigarette smoke. The effect on mice fed acrolein was dramatic: they suffered injuries to their gastrointestinal tracts as well as a whole-body response called “acute phase response,” a dramatic attempt by the body to avoid septic shock.† Markers of inflammation and other signs of acute infection also went up dramatically—sometimes by a hundredfold. Daniel J. Conklin, the cardiovascular physiologist who did this work, told me he was “stunned” to find that the dose required to provoke some version of this response was entirely possible from the levels of acrolein typically consumed on a daily basis, especially among people eating fried foods.
Aldehydes have not yet been officially classified as a toxin, but even so, there have been fewer experiments on humans to date.* One exception was a trial in New Zealand on diabetic patients. Those who were fed “thermally stressed” safflower oil had a significantly higher level of markers for oxidative stress than those consuming olive oil. In fact, olive oil has consistently been shown to produce fewer oxidation products than do polyunsaturated oils like soybean and corn. Olive oil, a monounsaturated fat, as you might remember, has only one double bond to react with oxygen, whereas vegetable oils are polyunsaturated, with many double bonds. However, the fats that produce the fewest oxidation products are those without any double bonds: the saturated fats found in tallow, suet, lard, coconut oil, and butter.
In 2008, Csallany presented her findings to her colleagues, mostly industry employees, at a meeting of the American Oil Chemists’ Society (AOCS) in Salt Lake City. “First they were alarmed and then nothing,” she said. And in London, a team of researchers have repeatedly tried to alert people of the problem through the news media and at professional conferences. The team wrote a letter to the journal Food Chemistry in 1999 entitled, “Warning: Thermally-Stressed Polyunsaturates Are Damaging to Health,” followed by a paper directed to “alert the foodservice industry” to health problems. Yet they too found little interest. Other researchers in the field are molecular biologists or biochemists, a world away from studying actual food items or making nutrition policy; as Rudolf Jörg Schaur, another of the HNE Club founders, wrote to me when I asked him if scientists were concerned about the increasing use of trans-free liquid oils in restaurants, “Since I am not a food chemist, I do not know.”
In 2006, the European Union formed a group of international researchers to understand better these lipid oxidation products and their implications for health. However, ADM’s Mark Matlock told me that there was nothing the industry could do about the production of aldehydes in their oils. Some restaurants were using specialized lowlinoleic or high-oleic oils, but regular oil (usually soybean or canola) was still the cheapest option. Kathleen Warner, an oil chemist who worked with the USDA for more than three decades and also directed the committee on heated oils for the AOCS for many years, told me that the best solution was simply to “hope” that restaurants filtered and changed their frying oils frequently and had good ventilation systems. Large fast-food chains also employ sophisticated techniques such as replacing the air over fryers with a “nitrogen blanket” and using micro-electric fields to minimize oxidation products. Warner confirmed that the aldehydes were “toxic,” however, and therefore a problem. Poli, the HNE Club co-founder, said he couldn’t understand why nutrition experts were so preoccupied with cholesterol, a vital molecule for many basic biological functions in the body, while ignoring HNE, a potential “killer” molecule. Another longtime oil chemist, Lars Wiedermann, who worked for many different food companies including Kraft and Swift & Co. from the early 1950s, told me that aldehydes and other toxic products need more mainstream attention: “Someone will surely discover how deadly used frying oils are,” he said.
Mark Matlock at ADM told me that the industry is waiting to see if the FDA takes an interest, since the FDA is the only agency that can formally designate something a “toxin.” So I asked to speak to scientists there. After months of delay, the FDA press office finally responded that while the agency was aware that oxidation products such as “alpha-beta unsaturated aldehydes” can form in heated polyunsaturated oils, there wasn’t yet enough information about their health effects. Is the agency working toward finding more information? Not yet. For now, it appears that the agency isn’t interested in knowing more about the oils that are a principal alternative to trans fats in baked and fried foods, billions of pounds of which are consumed by Americans each year.*
However, the FDA has been investigating other strange compounds that pop up in vegetable oils during processing: monochlorpropane diols and glycidol esters (MCPDs), which are also produced by heat and have been targeted by the European Food and Safety Authority for regulation due to their potential to cause cancer and kidney disease, among other things. Even though they occur only in trace amounts, Matlock told me that companies such as ADM are still working to get rid of them. Sound familiar? We are once more confronted by the unknown health consequences of vegetable oils, a century after they were first introduced into the United States.
From the earliest clinical trials in the 1940s, in which diets high in polyunsaturated fats were found to raise mortality from cancer, to these more recent “discoveries” that they contain highly toxic oxidation products, polyunsaturated oils have been problematic for health. They have nevertheless multiplied in use more than any other single foodstuff over the course of the twentieth century, fueled in large part by expert recommendations to eat more of them.
For more than sixty years, Americans have been told to eat polyunsaturated vegetable oils instead of saturated fats. This advice has been based on the simple reality that vegetable oils lower total cholesterol (and LDL-cholesterol, too, as later discovered). The fact that vegetable oils also create toxic oxidation products when heated and trigger inflammatory effects linked to heart disease, are, it seems, less important to mainstream nutrition experts, whose focus hasn’t wavered from cholesterol. Most Americans don’t realize that their nutritional advice is based on such a narrow set of health concerns, nor that large edible-oil companies have been contributing funds to their trusted, guiding institutions, such as the AHA, as well as to schools of medicine and public health. And while the scientists at large food manufacturers might understand the problems of unsaturated oils, they have not had alternatives to work with due to the prevailing stigma against saturated fats. Everyone has therefore gotten on board with the advice to use vegetable oils in both the home and industrial kitchens alike.
Our consumption has moved from saturated fats at the beginning of the twentieth century to partially hydrogenated oils to polyunsaturated oils. We have therefore unwittingly been subject to a chain of events starting with the elimination of animal fats and eventually winding up with aldehydes in our food. Looking ahead, it is little consolation that the FDA is poised to ban trans fats entirely, which will make liquid oils and their oxidation products even more common. Mom-and- pop restaurants, local cafeterias, and corner bakeries will then follow in the footsteps of the large fast-food restaurants in eliminating trans fats but will be less likely to employ rigorous oil-changing and ventilation standards in their operations. Despite the original good intentions behind getting rid of saturated fats, and the subsequent good intentions behind getting rid of trans fats, it seems that the reality, in terms of our health, has been that we’ve been repeatedly jumping from the frying pan into the fire.
The solution may be a return to stable, solid animal fats, like lard and butter, which don’t contain any mystery isomers or clog up cell membranes, as trans fats do, and don’t oxidize, as do liquid oils. Saturated fats, which also raise HDL-cholesterol, start to look like a rather good alternative from this perspective. If only saturated fats didn’t also raise LDL, the “bad” cholesterol, which remains the key piece of evidence against them. But like so many of the scientific “truths” that we believe but which upon examination start to crumble, maybe the LDL-raising effect isn’t quite an incontrovertible certainty, either.
*McDonald’s and Burger King list these oils as ingredients on their websites but would not confirm the cleaning problems.
†Even though people spend on average only 1.8 percent of their time in restaurants, they get about 11 percent of their exposure to tiny, potentially damaging airborne particles during this time, according to one analysis (Wallace and Ott 2011).
‡A team in Taiwan, which includes molecular biologists, toxicologists, and chemists, was formed due to concern about high rates of lung cancer among women living in Shanghai, Singapore, Hong Kong, and Taiwan. The team began investigating the possibility that heated cooking oils might be playing a role, since wok cooking with vegetable oils in unventilated space is common in Taiwan. (Some analyses show that in the United States, too, women who have never smoked have higher rates of lung cancer than do men) (Zhong et al. September 1999; Zhong et al. August 1999; Young et al. 2010).
**The unnatural oxidation products of heated oils are still being discovered. In addition to free radicals and aldehydes, these compounds include sterol derivatives, a plethora of products formed from degraded triglycerides, and other oxidized decomposition compounds. There are other unnatural chemical compounds, too, created by processes other than oxidation, including hydrolysis, isomerization, and polymerization (Zhang et al. 2012).
***The recommended frying temperature is 180 degrees Centigrade, but a study conducted by a leading biochemist found that restaurants almost always fry at higher temperatures (Firestone 1993).
††While the outward symptoms of the shock are few, significant changes take place inside the body, causing a dramatic increase in proinflammatory markers, a rise in some kinds of cholesterol, and a drop in serum total protein and albumin. ****Determination of a toxin is usually drawn from animal experiments. Human data may come from epidemiological studies, but epidemiologists have yet to study the issue of heated polyunsaturated oils in restaurant fryers, since usage only became common after the FDA enacted its labeling rule in 2006.
*****The day that the FDA proposed banning all trans fats in late 2013, partly in response to a petition by Fred Kummerow, he told me that he knew about the problem of oxidation products produced by heated polyunsaturated oils; in fact, he had done some of the original research on them himself in the 1950s. He said it was “unfortunate” that companies were now using regular oils for their frying operations and suggested that perhaps McDonald’s and Burger King could start broiling their french fries instead (Kummerow, interview with author, November 7, 2013).
This article appeared in Wise Traditions in Food, Farming and the Healing Arts, the quarterly journal of the Weston A. Price Foundation, Summer 2014