The World Health Organization threw the food and nutrition world into a tizzy yesterday with the publication of a Lancet article claiming that processed meat causes cancer and red meat is a likely carcinogen. The media did their very best to fuel the frenzy, with the LA Times publishing a piece exhibiting an abysmal lack of skepticism and possibly a gross misunderstanding of what the position statement actually means. The New York Times covered the story adeptly, but still failed to adequately address the critical issue of risk.
WHAT DOES IT TAKE TO BE CONSIDERED CARCINOGENIC?
A quick peruse of the WHO website and a link to Q & A provided by the International Agency for Research of Cancer (IARC) is cause for pause. The agency reviews everything from chemicals, environmental exposure, physical agents like radiation, biological agents like viruses, and personal habits.
A group of experts convenes to assess the research of a particular issue to determine if indeed it has the potential to cause cancer. So far the IARC has found that 400 of 900 different agents in the monograph series can be identified as carcinogenic, probably carcinogenic, or possibly carcinogenic to humans. Only one agent has not been found to present a hazard to humans, which may mean almost everything causes cancer or that the working groups are very good at selecting agents that are linked to cancer.
Still, the remarkable number of positive findings is not particularly surprising since many naturally occurring compounds in foods are known as to be toxic and likely to cause cancer in animals. What is appalling is that none of the initial reports I read clarified that IARC’s mission is to identify hazards, but purposefully does not address the issue of possible risk.
The Q & A paper provided by IARC makes these two specific points:
1. In 2014 the IARC Monographs Programme recommended evaluating the role of red meat and cancer based on epidemiological studies suggesting that small increases in the risk of several cancers may be associated with high consumption of red meat or processed meat.
2. IARC classifies processed meat as a carcinogen and red meat as a probably carcinogen based on the strength of evidence but does not assess the level of risk to meat eaters. (emphasis mine)
So what is the risk of eating processed meat? In the New York Times coverage, we are told that for every 2 ounces of processed meat intake per day, the risk of colorectal cancer rises by a factor of about 1.1 or 1.2. Is that significant? Consumers would have to know how to assess research to determine risk, leaving most vulnerable to media’s lack of clarity and penchant for fear mongering when it comes to nutrition and risk of anything.
THE LIMITS OF EPIDEMIOLOGICAL RESEARCH
The research is actually very weak linking processed meats to risk of cancer. Epidemiological studies cannot determine cause and effect. Peter Attia,MD, covers the basics in a brief post on his facebook page, keeping the facts short and simple:
The relative risk (RR) increase in cancer from smoking is 20x. At best, the RR increase from processed meat is 1.1 to 1.2x. So we’re talking about a 100- to 200-fold difference from smoking. Oh, and by the way, I’m pretty sure the great Bradford Hill said something about not being able to infer cause from epidemiology when the RR was less than about 4 or 5x. So…does consumption of processed meat lead to an increase in cancer risk? I don’t honestly know, but if it does, the risk is very small.
I have a paper re: meat’s place on the menu that will be published by Today’s Dietitian magazine in February and I learned plenty doing my research. The data re: processed meats is very messy and many researchers suggested publication bias, a dirty secret that isn’t often aired in public. All too often studies that find a negative association don’t get published or don’t get published in the most visible journals.
As I reviewed the literature I followed one particular epidemiological study that stated an increased risk of all cause mortality with eating processed meat. As I dug through the details in the discussion, the risk was only found for those who smoked, but not previous smokers or those who never smoked. That degree of detail didn’t make it into the abstract, and doesn’t gets reported in the popular press, either.
WHO STILL EATS LOTS OF MEAT?
Another critical issue regarding epiemiological research and red meat consumption spotlights the confounding impact from 40 years of telling people to not eat meat. Essentially we have created a bias that observational studies inadequately address. Too many people eating meat also smoke, drink excessively, don’t exercise and don’t eat fruits and vegetables, among a laundry list of unhealthful behaviors. How researchers supposedly tease the data to determine increased risk due to consuming a specific number of grams of processed meats per day mostly makes me want to roll my eyes.
In the IARC monograph, there is no discussion of dietary or other factors that could influence the potential carcinogenicity of eating meat. There is no comparison with the risk of eating fish or chicken because the risk from eating these foods wasn’t assessed. Neither is there a comparative assessment of eating a vegetarian diet. Basically there is no point of reference.
Lastly, there is no mention of the role of endocrine disruptors that bioaccumulate in the food chain. If there is any additional risk of cancer from eating meat, this could be it. But does this risk point to a problem with meat, or the utter failure on the part of the EPA and the FDA to regulate persistent organic pollutants and other endocrine disruptors that compromise metabolism and contribute to disease?
Essentially the monograph represents reductionist science at it’s best, focusing on the silo of risk from one food without perspective. Completely lacking in the body of epidemiological research is any meaningful address of how the animals are raised and their exposure to chemical agents. There is no discussion of the rest of the diet, nor the lifestyle habits of the consumers in question, but now we’ve opened yet another Pandora’s box.
PROBLEMS WITH HOW WE DETERMINE WHAT PEOPLE EAT
Even when diet is addressed in epidemiological research, scientists typically use the often criticized and problematic Food Frequency Questionnaire (FFQ). From my experience most people can’t remember what they ate yesterday much less specific details regarding the quantity and nature of foods consumed last month, last year, or 10 years ago. Most people don’t understand the composition of food well enough to answer the FFQ questions adequately. (I completed a survey just last month and even I ended up guessing at some of the more inane questions)
Some researchers blame the use of this tool for information that has directly influenced the misguided dietary advice for Americans over the last forty years– and call for it’s retirement.
ADDITIONAL CONCERN ABOUT INTERPRETING THE RESEARCH
In 1995, Gary Taubes penned a masterful piece exposing “the unholy alliance between epidemiologists, the journals and the press. When epidemiologists publish studies with weak links, no matter how cautious their findings, the media manages to blow them out of proportion. Taubes suggests that once the public links a behavior with risk, it is virtually impossible to discredit. Exactly what is happening here.
One observer at the of the working group reported that 7 of the 22 members of the scientific committee abstained from voting. I’d like to know more about that, and wonder if we ever will.
The purported risk of eating meat, even processed meat, fades significantly in the harsh light of scientific scrutiny. I am left wondering, who is doing the interpreting of these studies and what is their motivation? I’d like to see more assessment of how reducing meat in the diet stacks up with other dietary and lifestyle changes. We need some perspective here.