Is Everything We Know About Meat Consumption Wrong?

Stunning new recommendation says to keep eating it

by Kristen Monaco, Staff Writer, MedPage Today 

A new dietary guideline appearing in Annals of Internal Medicine declared eating red meat and processed meats won’t adversely affect health.

However, many experts in the field are hitting back stating there is simply no scientific evidence to support this claim made by Bradley Johnston, PhD, of Dalhousie University in Canada, and colleagues of NutriRECS.

Based upon five reviews, the 14-member voting panel concluded adults can continue to consume unprocessed red meat and processed meats, based on “low-certainty evidence” and yielding “weak recommendations.”

Systematic Reviews & Analyses

Johnston’s group assessed five systematic reviews, all simultaneously published in Annals of Internal Medicine — three of which also were meta-analyses. These three meta-analyses, which included cohort studies with at least 1,000 participants, looked at a few outcomes associated with red and/or processed meat consumption: risk of cancer, cardiometabolic, and all-cause mortality and incidence. The fourth paper included a review of 12 randomized clinical trials, which included a total of 54,000 participants. As for the fifth paper, this was a mixed-methods review not directly looking at health outcomes, but instead about the individual preferences and values of meat eaters.

Looking at health outcomes from unprocessed red meat consumption, the RCT review concluded with low- to very low-certainty evidence that eating red meat would have “little to no effect” on major health outcomes, such as diabetes, cancer, and heart events.

As for the three meta-analyses included, Johnston’s group noted that a reduction in unprocessed red meat consumption pointed to a significant, but “very small reduction” in cardiovascular mortality (risk difference of 4 fewer per 1,000 persons over 10.8 years), type 2 diabetes (6 fewer per 1,000 person over 10.8 years), and cancer mortality (7 fewer per 1,000 persons over lifetime), based on low- to very-low quality evidence.

These meta-analyses found similar outcomes when looking at consumption of processed meat. Based on low- to very-low quality of evidence, there was only a small risk reduction for cardiovascular mortality (4 fewer/1,000 persons in 10.8 years), type 2 diabetes (12 fewer/1,000 persons in 10.8 years), and cancer mortality (8 fewer/1,000 persons over lifetime).

Looking specifically at dietary preferences, that review concluded with low-certainty evidence that people who eat meat in general are unwilling to change their diet, even “in the face of undesirable health effects.”

A theme throughout the series of papers was that the quality of all evidence connecting dietary composition with health outcomes is generally low — typically based on self-report with infrequent updating of dietary habits over time.

In an editorial published with the study, Aaron Carroll, MD, MS, and Tiffany Doherty, PhD, of Indiana University’s Center for Pediatric and Adolescent Comparative Effectiveness Research, wrote that recommendation — however weak — to continue eating red and processed meat “is sure to be controversial, but it is based on the most comprehensive review of the evidence to date.”

They suggested pointedly that, precisely because the review is so comprehensive, “those who seek to dispute it will be hard pressed to find appropriate evidence with which to build an argument” without cherry-picking.

Instant pushback

But the group of studies that appear in Annals came under attack even before it was published.

In response, 13 researchers prominent in the diet-and-health field — including Harvard’s Walter Willett, MD, DrPH; David Katz, MD, of the True Health Initiative; Dean Ornish, MD, developer of the popular eponymous diet that recommends no meat; Kim Williams, MD, former American College of Cardiology president; and former U.S. Surgeon General Richard Carmona, MD — wrote to Annals editor-in-chief Christine Laine, MD, MPH, on Sept. 26 to request a preemptive retraction of the papers “pending further review.”

Even an author who helped with one of the meta-analyses signed onto the letter. Nevertheless, the request was denied.

Their letter argued that low-certainty evidence “is in no way a logical or even rational basis to recommend. If uncertain about what the data DO show, how much more so about what they do NOT?”

“This is, simply, an overt misrepresentation,” the letter continued. “Such distortion is a direct threat to public understanding, and public health.”

In another rebuttal to this guideline, the Department of Nutrition at the Harvard T.H. Chan School of Public Health, where Willett serves as chairman, issued a statement calling out the self-appointed panel on this guideline, saying how “[i]t’s unprecedented that a prestigious medical journal like Annals would publish guidelines by a self-appointed panel.”

“Would it publish clinical guidelines on other controversial topics such as PSA or mammogram screening developed by self-appointed groups in the future?” their statement continued. “It is one thing to publish a study that challenges the existing paradigm, but another to publish controversial guidelines that contradict the evidence.”

The American Institute for Cancer Research, along with a slew of other international health organizations including the American Society for Preventive Oncology and World Cancer Research Fund, also issued a response statement asking the public to dismiss this guidance and continue to limit their red, processed meat intake.

Limitations

In reviewing this evidence, researchers writing the Annals papers relied on GRADE criteria. In an interview with MedPage Today, Christopher Gardner, PhD, of Stanford University, and co-signee of the response letter to the publication, explained that a “huge problem” with this assessment is how GRADE criteria is only suitable for drug studies, not nutrition studies.

Gardner also pointed out how only one of the panel’s five reviews included randomized trials, while the rest relied on observational epidemiologic data. Calling observational epidemiology “outdated,” he noted that this type of nutrition assessment misses the important question: “if they eat less meat, what do they eat more of … the health benefit of less meat is contingent on ‘what instead?’ which they don’t address.”

As for the one review including randomized trials, many of the trials were funded by the meat industry, and none looked solely at health outcomes from red and/or processed meats. Instead, most studies looked at outcomes of the Mediterranean diet. On the other hand, by far the largest trial included in this set and that drove the overall results was the government-sponsored Women’s Health Initiative.

“It is astonishing that the Lyon Heart trial was excluded because the results were ‘too good,'” Frank Hu, MD, PhD, of the Harvard T.H. Chan School of Public Health, and fellow co-signee of the response letter to the publication, told MedPage Today. “It’s surprising that other trials like PREDIMED [later retracted and republished with revisions], DPP, etc. were not included. A reduction in red meat was also included in these trials, which showed positive results.” (In PREDIMED, though, red meat was discouraged in both diets and more strongly for the control vs Mediterranean diet.)

This sentiment was echoed by Gardner, who called the trials included in the Annals review “horrible” and “not relevant.” One of the trials included was the Women’s Health Initiative (WHI), which looked at outcomes of a low-fat diet. In the response statement from Harvard T.H. Chan School of Public Health’s department of nutrition, one WHI investigator called this a “mis-use and misinterpretation of data from the WHI.”

The Harvard statement also pointed out that there has never actually been a randomized controlled trial looking at long-term health outcomes from red meat reduction due to “practical and ethical reasons” — a point also emphasized by the Annals authors.

In one of the three meta-analyses used for the review, there were contradicting conclusions on the health outcomes assessed. This analysis by Vernooij and co-authors found eating less red and processed meat was tied to a statistically significant decrease for the risk of all these outcomes:

  • Total all-cause mortality: relative risk 0.87 (95% CI 0.82-0.92)
  • Cardiovascular mortality: RR 0.86 (95% CI 0.79-0.94)
  • Cancer mortality: RR 0.89 (95% CI 0.83-0.96)
  • Type 2 diabetes: RR 0.76 (95% CI 0.68-0.86)

Hu pointed out how “the consistent findings and narrow 95% confidence intervals actually reflect the certainty of the data, contrary to the authors’ own assertions about low certainty of the data,” and that few lifestyle or nutritional exposures are tied to this magnitude of risk reduction.

“These risk reductions are not small from a public health point of view and are stronger or comparable to other public health problems such as low consumption of fruits and vegetables, physical inactivity, passive smoking, air pollution, etc. which were based on similar types of prospective cohort data because it is not feasible or ethical to conduct large long-term trials on these exposures,” he added.

One of the co-authors on this particular meta-analysis, John Sievenpiper, MD, PhD, of the University of Toronto, who also signed off on the response letter to Annals of Internal Medicine, explained in a statement that “[u]nfortunately, the leadership of the paper chose to play up the low certainty of evidence by GRADE, as opposed to the protective associations that directly support current recommendations to lower meat intake.”

“Very few nutritional exposures are able to show associated benefits on the big three of all-cause, cardiovascular, and cancer mortality as well as type 2 diabetes. The signals would be even stronger if one considered substitution analyses with plant protein sources or investigated dose-response gradients which are used to upgrade data by GRADE, both of which I had requested. Unfortunately, I never saw the galley proofs to ensure that these changes had been made,” he concluded.

Ultimately, many of these experts underscored the importance of sticking to current dietary guidelines already in place, recommending a healthy, balanced diet — limiting red and processed meat intake — in order to help prevent chronic diseases.

For their part, Carroll and Doherty noted that continuing to beat the meat-is-bad drum is probably pointless. “We have saturated the market with warnings about the dangers of red meat,” they wrote. “It would be hard to find someone who doesn’t ‘know’ that experts think we should all eat less. Continuing to broadcast that fact, with more and more shaky studies touting potential small relative risks, is not changing anyone’s mind.”

Perhaps more fruitful would be to undertake “a major overhaul of the methods for communicating nutritional data in ways that might get through to target populations and change health outcomes,” the editorials wrote. For example, “ethical concerns about animal welfare” and the environmental impacts of meat production might be more persuasive, they suggested.

“[T]hey have the added benefit of empirical evidence behind them,” Carroll and Doherty noted dryly.

Members of the voting panel declared they had no financial or intellectual conflicts of interest. Other authors did report disclosures.

Carroll reported publishing a book on nutrition and science (as have many of those criticizing the papers) for which he receives royalties. Doherty declared having no relevant financial interests.

LAST UPDATED 

Comments Are Closed