publication date: Oct. 4, 2019
Don’t put red, processed meat back on the menu:
Cancer and dietary experts dispute controversial directive
By Matthew Bin Han Ong and Alex Carolan
Global health organizations, federal health agencies, and cancer epidemiology experts say they aren’t swayed by just-published recommendations on consumption of red and processed meat.
An international group of researchers earlier this week published five systematic reviews of evidence and concluded that it’s okay for adults to continue to eat unprocessed red meat and processed meat at their average consumption levels.
The recommendation, made by the panel of 14 researchers from seven countries and published in Annals of Internal Medicine Sept. 30, runs contrary to almost all other existing guidelines.
“Among 12 randomized trials enrolling about 54,000 individuals, the researchers did not find statistically significant or an important association between meat consumption and the risk of heart disease, diabetes, or cancer,” a statement from the Annals of Internal Medicine reads. “Amongst cohort studies following millions of participants, the researchers did find a very small reduction in risk amongst those who consumed three fewer servings of red or processed meat per week. However, the association was very uncertain.”
The lead author of the recommendation paper, Bradley Johnston—who is also a co-founder of the guidelinemaking group, NutriRECS—has previously received funding from the International Life Sciences Institute to conduct a 2016 study on consumption of sugar. ILSI is an industry trade group largely supported by pharmaceutical companies and food companies, including Cargill, one of the largest beef processors in North America.
After news of this conflict of interest surfaced, Johnston argued that his past relationship with ILSI wasn’t subject to disclosure and didn’t affect the group’s recommendations on red and processed meat. The COI subplot of this complex story is developing.
Leading organizations in public health and oncology say the authors’ interpretation of findings should not be used to inform public dietary guidelines, because these new recommendations are “not supported by the scientific evidence.”
In six interviews with The Cancer Letter, epidemiologists, cardiometabolic scientists, and experts in guidelinemaking characterize the recommendations as “confusing,” “sensational,” “self-appointed,” “so-called guidelines” that are a “disservice to the public.”
“Leading global cancer experts do not agree with these authors’ interpretation of the scientific evidence,” said Jill Reedy, chief of the Risk Factor Assessment Branch, Epidemiology and Genomics Research Program at the NCI Division of Cancer Control and Population Sciences. “[These] cancer experts continue to recommend the guidance that existed before, regarding limiting red meat intake and eating little, if any, processed meat for cancer prevention.”
The full transcript of these conversations—with physicians and researchers at NCI, the National Heart, Lung, and Blood Institute, the University of California, San Francisco, MD Anderson Cancer Center, the American Cancer Society, and The University of North Carolina at Chapel Hill—appears here.
These experts, who independently reviewed the studies for The Cancer Letter, concluded that:
The recommendations published in Annals of Internal Medicine run counter to the totality and preponderance of evidence, which demonstrates that limiting or reducing consumption of red and processed meats is important for prevention of cancer and cardiometabolic diseases,
The systematic reviews confirm existing evidence that there is an association between consumption of red and processed meats with adverse health outcomes, and that
The investigators should be saying, “We don’t know,” instead of “Keep going,” because the “low to very low” certainty of evidence provided by the systematic reviews is “insufficient” for meeting the threshold for changing or informing dietary guidelines.
“No need to reduce red or processed meat consumption”
Leading up to the publication of the controversial studies Monday evening, reporters around the world received an email with seven PDFs, and an embargoed press release with the headline:
“New guidelines: No need to reduce red or processed meat consumption for good health.”
“Based on a series of five high-quality systematic reviews of the relationship between meat consumption and health, a panel of experts recommends that most people can continue to consume red meat and processed meat at their average current consumption levels,” the press release from Annals of Internal Medicine said.
In summary, the systematic reviews concluded that:
Low to very-low-certainty evidence suggests that diets restricted in red meat may have little or no effect on major cardiometabolic outcomes and cancer mortality and incidence.
The possible absolute effects of red and processed meat consumption on cancer mortality and incidence are very small, and the certainty of evidence is low to very low.
The magnitude of association between red and processed meat consumption and all-cause mortality and adverse cardiometabolic outcomes is very small, and the evidence is of low certainty.
Low or very-low-certainty evidence suggests that dietary patterns with less red and processed meat intake may result in very small reductions in adverse cardiometabolic and cancer outcomes.
Low-certainty evidence suggests that omnivores are attached to meat and are unwilling to change this behavior when faced with potentially undesirable health effects.
The authors derived these conclusions based on:
A systematic review of 12 randomized trials comparing diets lower in red meat with diets higher in red meat, to summarize the effect of these diets on the incidence of cardiometabolic and cancer outcomes in adults,
Three systematic reviews of 23 cohort studies evaluating the relationship between red meat and processed meat with cancer mortality and incidence, all-cause mortality, cardiometabolic outcomes, quality of life, and satisfaction with diet, and
A mixed-methods systematic review of values and preferences regarding meat consumption.
The publication packet also included:
An accompanying editorial from two researchers at the Indiana University School of Medicine endorsing the studies as “the most comprehensive review of the evidence to date,” and
A paper that states the group’s recommendation and describes their guideline development process.
The final document, titled “Unprocessed Red Meat and Processed Meat Consumption: Dietary Guideline Recommendations,” and labeled “CLINICAL GUIDELINE” in the journal, was developed through a guideline development process created by NutriRECS.
What is NutriRECS?
NutriRECS, or Nutritional Recommendations and Accessible Evidence Summaries Composed of Systematic Reviews, bills itself as “an independent group” of researchers founded by three researchers—an epidemiologist from Canada, a health services researcher from Spain, and another epidemiologist from Poland.
“NutriRECS is an independent group with clinical, nutritional and public health content expertise, skilled in the methodology of systematic reviews and practice guidelines who are unencumbered by institutional constraints and conflicts of interest, aiming to produce trustworthy nutritional guideline recommendations based on the values, attitudes and preferences of patients and community members,” its website states.
NutriRECS receives funding from members’ universities who pay salaries to complete the work, and occasional grants that pay graduate student salaries, said NutriRECS member Gordon Guyatt, Distinguished Professor in the Department of Clinical Epidemiology and Biostatistics at McMaster University, who is an author on all five systematic reviews as well as the recommendations document.
“Our business model is the usual academic model—get your work done by hook or by crook, volunteer labor of folks working for an education and their names on prominent papers, graduate students whose work will be part of their dissertations, and visiting scholars and professors who come to learn how we do things and how to get publications in top journals,” Guyatt said to The Cancer Letter.
Several members are also involved in the Cochrane Collaboration of meta-analysts.
A conversation with Guyatt appears here.
A NutriRECS co-founder, Johnston, is intimately involved in the design and development of the studies on red and processed meats published in the Annals of Internal Medicine. Johnston is lead author of the recommendation paper, an author on all five systematic reviews, and a corresponding author on four of the systematic reviews.
“One of the gentlemen who did a postdoctoral fellowship with me, Brad Johnston, now works out of Halifax, and I continue to collaborate with him,” Guyatt said. “He has taken a career direction where he wants to focus on nutrition, and particularly on nutritional guidelines, having noted that many nutritional guidelines are not high quality, and are quite flawed, and trust me, the guidelines are necessary.
“We are currently planning our next project, which has to do with fats.”
Sugar, and previous COIs
Though Johnston and his fellow authors of the studies on red and processed meat reported no conflicts of interests, Johnston has received funding from food companies in a December 2016 study on sugar consumption.
That study, “The Scientific Basis of Guideline Recommendations on Sugar Intake: A Systematic Review,” also published in Annals of Internal Medicine, reached similarly formulated conclusions.
Johnston and his team conducted a systematic review of nine public health guidelines on sugar recommendations—each suggesting decreasing consumption of foods containing nonintrinsic sugars—including those by the Dietary Guidelines for Americans and the World Health Organization, and concluded that the “quality of evidence supporting [these] recommendations was low to very low.
“Guidelines on dietary sugar do not meet criteria for trustworthy recommendations and are based on low-quality evidence,” the study concluded. “Public health officials (when promulgating these recommendations) and their public audience (when considering dietary behavior) should be aware of these limitations.”
The “low to very low” characterization of the evidence is similarly applied to Johnston and his team’s conclusions on consumption of red and processed meats.
“According to their conflicts of interest, they have none [currently],” Christopher Gardner, director of nutrition studies and Rehnborg Farquhar Professor of Medicine at the Stanford Prevention Research Center, said in an Oct. 3 interview on KQED San Francisco, NPR’s largest member station. “Two years ago, this same group said there’s no clear evidence that sugar is bad for us, so we should keep eating sugar at the same amount that we are right now.
“That particular one was funded by ILSI, which is the International Life Science Institute, and that does have some connections to Big Food.
“I haven’t seen big connections to cattle, but in the past this group has done that. I think this one, they tried really hard to be squeaky clean. It’s hard to say for sure,” Gardner said. “So many authors, but they did present clean conflicts of interest statements to the journal.”
Johnston’s 2016 study on sugar was funded by the Technical Committee on Dietary Carbohydrates of ILSI North America. The authors wrote the protocol—the scope of which was reviewed and approved by ILSI—and conducted the study independently from ILSI.
At the time, Johnston was a methods consultant to ILSI, held investigator-initiated grants from BioK+ and Genzyme, and was funded by a joint grant from Nestlé and MITACS Accelerate.
ILSI was created by a Coca-Cola executive in 1978. Members of ILSI have included food companies McDonald’s, General Mills, Nestlé, PepsiCo, and Cargill, as well as pharmaceutical companies Pfizer and Sanofi.
“The only person with overlap in the two projects was Brad Johnston, who has not taken anything that would constitute a conflict of interest in this one,” Guyatt said. “I don’t know when Brad took the money from this group that has all sorts of food makers in it.
“I think Brad reported his conflict of interest. I think Brad told me that there was somebody in his group who did not [report their COIs on the sugar study].”
One of Johnston’s co-authors in the sugar study, Joanne Slavin, had received laboratory funding from entities that include the Minnesota Beef Council, Danone, and Coca-Cola. Slavin doesn’t figure as an author on the red and processed meat studies.
A correction added by the Annals of Internal Medicine to the 2016 sugar study states:
“Disclosures from two authors (Drs. Johnston and Slavin) were explained in the Disclosures section of the article. The role of the funding source, ILSI, was also clarified in the article and in the Financial Support section.”
The correction was added after the Associated Press obtained emails showing that ILSI had “reviewed” and “approved” the study’s protocol, according to The New York Times.
“If Brad has managed to get some funding specific to NutriRECS, I haven’t heard,” Guyatt said to The Cancer Letter.
Johnston did not respond to emails from The Cancer Letter.
“That money was from 2015 so it was outside of the three-year period for disclosing competing interests,” Johnston said to The New York Times Oct. 4. “I have no relationship with them whatsoever.”
The International Committee of Medical Journal Editors recommends that investigators report “all sources of revenue paid (or promised to be paid) directly to you or your institution on your behalf over the 36 months prior to submission of the work. If there is any question, it is usually better to disclose a relationship than not to do so.”
The Annals of Internal Medicine relies on the ICMJE conflict of interest guidelines.
Johnston should have disclosed his past relationship with ILSI, especially because of the gravity of his group’s recommendations on meat consumption, said Arthur Caplan, the Drs. William F. and Virginia Connolly Mitty Professor of Bioethics at New York University Langone Health and the founding director of the Division of Medical Ethics.
“If you’re going to try publishing something like ‘smoking doesn’t cause cancer, it’s actually good for you,’ I’d probably try to disclose about 30 years of my past funding of financial relationships,” Caplan said to The Cancer Letter. “And meat is very similar.”
In the red and processed meat studies, researchers on the paper disclosed whether they eat red and processed meat and how often they eat it.
“Trustworthy guidelines demand minimizing conflict of interest,” Guyatt said. “So, nobody on our panel had any financial conflict of interest, and we minimized intellectual conflicts of interest. And since some people might perceive what you’re eating as a conflict of interest we declared, as you have noted, the meat in our diet.”
Guyatt said he is a pescetarian, for animal welfare and environmental reasons. The researchers did not include these variables in their guideline development process.
“Considerations of environmental impact or animal welfare did not bear on the recommendations,” the guidelines document states.
Cancer research groups are unconvinced
The NutriRECS recommendations “undermine public confidence in dietary advice,” said Nigel Brockton, vice president of research at the American Institute of Cancer Research.
“We stand by the rigor of our research methodology and our Cancer Prevention Recommendation that people should limit red meat intake to less than 12-18 oz per week and avoid processed meat,” Brockton said in a statement Sept. 30. “The underlying results reported by the NutriRECs group are actually consistent with this advice, but they dismiss these results based on the limitations of some contributing research methods.”
The best available evidence supports an increased cancer risk, AICR said in a statement.
“The NutriRECS research results are not significantly different from what World Cancer Research Fund/American Institute for Cancer Research’s 2018 report found, and indeed seem to verify WCRF/AICR’s findings,” AICR said in a statement. “However, the NutriRECS researchers have made what is a confusing interpretation of the results, which has led to this unnecessary recommendation to the public.
“The NutriRECS recommendation does not separate out red and processed meat, and this suggests that three or four portions of processed meat a week do not affect cancer risk significantly enough to warrant a reduction in the amount people eat. This conclusion from NutriRECS is not supported by the scientific evidence.”
The public should continue to use the Dietary Guidelines for Americans, said Holly Nicastro, program director in the Division of Cardiovascular Sciences at NHLBI.
“Dietary guidelines in the United States draw from the same body of evidence that these researchers of the meta-analysis had available to them,” Nicastro said to The Cancer Letter. “The difference is that the [Dietary Guidelines for Americans] look at dietary patterns or eating patterns as a whole and conclude that a reduction in red and processed meat would be beneficial. The authors of this analysis looked at red and processed meat in isolation.”
The NutriRECS group weighted randomized trials more heavily than observational studies, which influenced the conclusions of the systematic reviews, critics say.
“Easiest way to think of this report is, a group of people who have expertise in epidemiology got together as a grand jury and they decided to review literature on red meat and processed meat and cancer,” said Otis Brawley, Bloomberg Distinguished Professor of Oncology and Epidemiology and associate director for community outreach and engagement at the Bloomberg School of Public Health and Johns Hopkins Kimmel Cancer Center. “And when you do this sort of thing, you can introduce biases. They openly admit that they down-weighted the importance of non-randomized trials, and they considered randomized trials to be the gold standard.
“And by down-weighting the importance of non-randomized trials, they got this result,” Brawley said to The Cancer Letter. “Now, many people in the community think, for a number of different reasons, that you really can’t just disparage non-randomized trials when you’re asking a question about long-term behavior, like what have you been doing over a long period of time.
“They didn’t exclude non-randomized trials, they actually just mathematically said these are far less important than randomized trials in making their decision. These folks came out with the finding that they found because they discounted non-randomized trials, and I don’t think you can discount non-randomized trials because of behavioral considerations.”
The NutriRECS authors have created a “big splash” that is leading to a lot of confusion, said NCI’s Reedy.
“We see that these new so-called guidelines aren’t justified, but keep in mind, these so-called guidelines are also contradicting the evidence that was generated from these authors’ own meta-analyses,” Reedy said. “This was a self-appointed panel, and I know that we’ve heard from other reports that there is some disagreement among the authors. Not all the authors agreed with the language and the final papers.”
Lifestyle factors, including diet, are difficult to study in randomized clinical trials, which would require tracking eating habits over many years, said Carrie Daniel-MacDougall, associate professor of epidemiology and director of the Bionutrition Research Core at MD Anderson.
“The kind of trial that people want is impossible,” Daniel said to The Cancer Letter. “We cannot get individuals to leave everything in their diets the same and just change meat intake and follow them for 10 years to see who gets cancer.
“I’ve worked on studies where we’ve actually found fairly large effect sizes from eating red and processed meat with large prospective studies. It’s individual to the cancer. In colorectal cancer, we’ve seen very large effect sizes. In breast and prostate cancer, we see lower effect sizes, because they’re totally different cancers that develop from totally different mechanisms.
“So, if you mash it all together, the low ones and the high ones, you’re going to get something modest. But that doesn’t mean that red and processed meat doesn’t cause colorectal cancer. From a mechanistic standpoint, it does.”
Public health organizations have accepted that cigarettes are a cause for cancer without needing proof from randomized studies, said Rita Redberg, professor of medicine and a cardiologist at the University of California, San Francisco.
“This reminds me of when the tobacco industry said, ‘No, smoking hasn’t been proven to cause cancer, because there were no randomized studies,’” Redberg, who is also editor of JAMA Internal Medicine, said to The Cancer Letter. “I mean, you just don’t need randomized studies when you’re talking about big lifestyle issues like smoking or like food.”
To rate the certainty of evidence, the NutriRECS researchers used GRADE (Grading of Recommendations, Assessment, Development and Evaluations), a system typically used in drug trials. Under this grading system, the researchers rated the quality of observational studies as weak or very weak, said Marji McCullough, senior scientific director of epidemiology research at the American Cancer Society.
“In reviewing the evidence for diet and lifestyle, we tend to use different review criteria, because it’s really difficult to do long-term trials of diet and cancer,” McCullough said to The Cancer Letter. “They would downgrade the evidence if there weren’t repeated measures of diet during the course of a prospective analysis. And they would downgrade if, for example, family history was not included in the model.
“The American Cancer Society continues to recommend that people limit their consumption of unprocessed red meat, and especially processed meat, based on the totality of the existing evidence and conclusions by the World Health Organization that processed meats are carcinogenic and unprocessed red meat is considered a probable carcinogen.”
Evidence-based medicine based on low-certainty data
Because Guyatt, Johnston, and their team found “low to very low” certainty of evidence, and the magnitude of effect of red and processed meat consumption on cardiometabolic and cancer outcomes is small, are the recommendations to maintain consumption levels reasonable?
Evidence-based medicine is the practice of using the best available evidence to make decisions, according to Guyatt, who says he coined the phrase “evidence-based medicine.”
“Well, you have to make choices; right?” Guyatt said. “We don’t say, ‘No, sorry. No randomized trials, can’t help you.’ Instead, we use the best evidence available.
“This is certainly not ideal to have only low-quality evidence, but people need to make decisions about their red meat consumption, and the decisions should be based on the best evidence available.”
Russell Harris, emeritus professor of the Public Health Leadership Program at The University of North Carolina at Chapel Hill, disagrees.
“What [NutriRECS] could have said is, ‘The evidence is insufficient,’” Harris said to The Cancer Letter. “They could say, ‘This evidence is so lousy that we can’t tell you whether eating meat is bad. Certainly, there’s no signal here that it’s good for you, by the way. But we can’t tell you whether it’s bad for you or not. So, you’re going to have to decide this based on other things.’”
Harris is a former member of the United States Preventive Services Task Force, an independent, volunteer panel of national experts in disease prevention and evidence-based medicine. The Centers for Medicare and Medicaid Services use USPSTF recommendations to inform coverage.
“Let me point out, that’s the reason the [Task Force] has what they call an ‘I,’” Harris said. “They’re one of the only guideline groups that have something called an ‘I,’ insufficient evidence—that means we looked at the evidence as hard as we could.
“And guess what? This doesn’t answer the question that we wanted to ask.”
The NutriRECS authors may be “calling out” others who have made strong recommendations on the consumption of red and processed meat, Harris said.
“To me, they made the same mistake that they’re calling others out for having made. That’s an error,” Harris said. “Is the evidence enough to make a recommendation? That’s really where we are with this. Is there enough evidence here to say anything other than, ‘I don’t know?’”
Daniel and Redberg said they continue to recommend a Mediterranean-style diet.
“I think something in the paper that was sort of buried was the dietary patterns that are traditionally lower in red and processed meats, like the Mediterranean diet,” Daniel said. “The Mediterranean diet has been associated—in several studies and trials—with lower risk of cancer and cardiovascular disease.
“Overall, eat a variety of foods, but not too much. That has not changed just because a paper has come out on this.”
“I like what Michael Pollan said, ‘Eat food. Not too much. Mostly plants,’” Redberg said. “And get regular exercise.”