Journal Advice to Eat Cancer-Causing Meats: Science or Clickbait?

In response, the journal has come under unprecedented fire by a broad group of health experts who have pointed out major flaws in the published articles and asked the journal to retract them. 

Background

Processed meat (bacon, sausage, hot dogs, ham, etc.) was classified in 2015 by the World Health Organization as a group 1 carcinogen. Evidence of cancer risk is clear and convincing for colorectal cancer, and the WHO report also observed associations between red and processed meat consumption and stomach, pancreatic, and prostate cancers.¹ Processed meat is also associated with breast cancer risk, among other health problems.² In 2017, the American Medical Association called for the elimination of processed meat from hospital menus,³ as did the American College of Cardiology.

Also relevant as background is the fact that the Annals of Internal Medicine has a history of publishing sensational articles on nutrition, notably the 2014 article⁴ that led Time magazine to put a swirl of butter on its cover with the headline “Eat Butter.” The 2014 Annals article in question seemed to exonerate “bad” (saturated) fat, but was shown to have major errors that forced the Annals to issue a correction.

In 2017, the Annals published an article seeming to exonerate sugar, claiming that evidence of harm was weak.⁵ The research was paid for by the International Life Sciences Institute (ILSI), an organization funded by food and agrochemical companies. The senior author and journal editor for this article were the same as for the new set of articles on meat.

In 2019, the Annals published a commentary concluding that nutrition modification, either by diet changes or supplements, was not justified in treatment guidelines for heart patients.⁶

Who wrote the articles and what did they say?

Bradley Johnston, working with a private organization called NutriRECS, conducted meta-analyses of prior studies that showed that reductions in red and processed meat are associated with reductions in cancer and cardiovascular risk. The articles overlap considerably in their methods and findings, but here is a summary of what they found:

Observational studies on cardiometabolic outcomes and cancer (Vernooji 2019).⁷ Summarizing the results of observational studies, the researchers reported that diets that are low in red and processed meat (compared with meatier diets) were associated with a 13 percent lower risk of all-cause mortality, 14 percent reduction in cardiovascular mortality, 14 percent reduction in nonfatal stroke, 24 percent reduction in type 2 diabetes, 10 percent reduction in overall cancer incidence, and 11 percent reduction in cancer mortality. Some findings were more striking, including a 59 percent reduction in extrahepatic cancer incidence, 64 percent reduction in gallbladder cancer incidence, and 56 percent reduction in pancreatic cancer mortality. All of these differences were statistically significant, meaning they were unlikely to be due to chance.

Observational studies on mortality and cardiometabolic outcomes (Zeraatkar, 2019a).⁸ Summarizing past studies on populations that differed in their meat consumption, the researchers reported that reducing red meat intake by three servings per week would cut cardiovascular mortality by 10 percent, stroke by 6 percent, myocardial infarction by 7 percent, and type 2 diabetes by 10 percent, all of which were statistically significant.

Similarly, reducing processed meat by the same amount would cut cardiovascular mortality by 10 percent, stroke by 6 percent, myocardial infarction by 6 percent, and type 2 diabetes by 12 percent. Again, all of these were statistically significant. 

Observational studies on cancer mortality and incidence (Ah Han 2019).⁹ A meta-analysis of observational studies showed that a reduction of three servings of meat per week would be expected to reduce overall cancer mortality by 7 percent. A similar reduction in processed meat would be expected to lead to reductions in esophageal (30 percent), colorectal (7 percent), and breast (10 percent) cancer incidence and an 8 percent reduction in overall cancer mortality. Each of these was statistically significant. 

Randomized clinical trials (Zeraatkar 2019b).¹⁰ The researchers summarized the results of only one randomized trial, the Women’s Health Initiative, a study designed to test the effects of modest diet changes, rather than the effects of reducing meat consumption. Nonetheless, participants did reportedly reduce meat intake slightly. While the authors of the Annals paper discounted the results of the WHI, in fact, after 19.6 years of follow-up, the intervention group had a 16 percent reduction for breast cancer followed by death and a 13 percent reduction in diabetes requiring insulin, both of which were statistically significant.¹¹  

Values and preferences (Valli 2019).¹² The researchers identified articles showing that omnivores enjoy eating meat and that many people feel, rightly or wrongly, that meat is important to health. It did not address whether these attitudes can be influenced by nutrition education, advertising, or changes in product availability, such as the recent rise in availability of meat alternatives.

Dietary guideline recommendations (Johnston 2019).¹³ Johnston concluded that the benefits of reducing meat consumption that had been identified were small and uncertain, and that many individuals are reluctant to reduce their meat consumption, despite their risks, writing:

 “…the panel believed that for the majority of individuals, the desirable effects (a potential lowered risk for cancer and cardiometabolic outcomes) associated with reducing meat consumption probably do not outweigh the undesirable effects (impact on quality of life, burden of modifying cultural and personal meal preparation and eating habits).”

Johnston did not take this argument to mean that there was little point in encouraging individuals to reduce their meat intake. Instead, he offered guidelines encouraging the continued consumption of these products. Specifically, Johnston, et al, wrote:

“The panel suggests that adults continue current unprocessed red meat consumption (weak recommendation, low-certainty evidence). Similarly, the panel suggests adults continue current processed meat consumption (weak recommendation, low-certainty evidence).”

The panelists are not identified in the report. The paper notes that this recommendation was not unanimous. Of the 14 panelists selected by Johnston to opine on his position, three dissented, calling for reducing meat intake.

In its publicity materials, the journal reinforced this message. In advance of publication, it issued a press release with the headline “New guidelines: No need to reduce red or processed meat consumption for good health.”

Who is criticizing them?

Prior to publication, the Annals editor, Christine Laine, was contacted by Neal Barnard, MD, FACC, of the Physicians Committee for Responsible Medicine, who also serves on the adjunct faculty of the George Washington University, David L. Katz, MD, MPH, of Yale University, and Walter C. Willett, MD, DrPH, of the Harvard School of Public Health, who pointed out that the journal’s press release would garner widespread media attention and promote the false notion that reducing red and processed meat intake does not lead to health benefits. The editor responded, agreeing that the press release was misleading and needed to be changed, but did not distribute a corrected release to the press.

The True Health Initiative, an organization including more than 500 health experts, asked the Annals to pre-emptively retract publication of these papers, because the conclusions were not supported by their findings.

One of the Annals article’s authors, John Sievenpiper, of the University of Toronto, called on the Annals to stop publication. Although he had participated in the analysis, he believed the conclusions were mistaken and joined the True Health Initiative’s appeal to the editor.  

Dr. Sievenpiper wrote:  

As a co-author of one of papers and a physician, I share the concern that the public may view this series of systematic reviews and meta-analyses and resulting clinical practice guideline as commissioned or endorsed by the American College of Physicians. Our review of dietary patterns low in red and processed meat showed associated reductions in the most important outcomes to patients and public health: all-cause mortality, cardiovascular mortality, and cancer mortality. Despite this finding, the recommendation from the guidelines paper (of which I was neither a part nor was I invited to review as a co-author of one of the reviews) was that adults continue consuming red and processed meat. I completely oppose this recommendation and worry about the lasting damage to public and planetary health.

What’s wrong with the Annals articles?

Apart from the fact that the recommended dietary guidance was contrary to the identified benefits of reducing meat consumption, the articles had several methodologic problems. They left out key data, used inappropriate analyses, and allowed their own unverified guesses about public resistance to diet changes to influence their recommendations.

Missing data. Numerous key studies were excluded from the NutriRECS analysis. The PREDIMED (Prevención con Dieta Mediterránea) study was the large and well-known study that established the ability of a Mediterranean diet that replaced red meat with more healthful foods to reduce cardiovascular risk.¹⁴ Those whose diets tended most toward vegetarian patterns had the largest reductions in all-cause and cardiovascular mortality.¹⁵

The DASH (Dietary Approaches to Stop Hypertension) study is the classic study showing that diet changes reduce blood pressure. The study’s dietary program specifically shifted the diet away from red meat toward more healthful choices.¹⁶

The Lifestyle Heart Trial showed that a low-fat vegetarian diet, as part of an overall healthy lifestyle, could reverse the progression of even severe coronary heart disease.¹⁷ Many other randomized clinical trials—summarized in meta-analyses—have shown that replacing meat and other animal products with healthier choices consistently improves blood cholesterol,¹⁸ body weight,2 blood pressure,¹⁹ and blood sugar control.²⁰

Problems with analyses. In conducting its meta-analyses, the Annals articles used only the most heavily adjusted findings from the studies that were reviewed. For example, a study assessing meat’s association with heart attack risk might be adjusted for cholesterol levels, overweight, and blood pressure. However, if meat causes heart problems because it increases cholesterol, body weight, and blood pressure, adjustment for these factors could cause meat’s deleterious effects to be no longer noticeable in reported statistics. While statistical adjustments are often necessary and appropriate, if used injudiciously they obscure real findings.

Guessing about reluctance to change. Perhaps the most controversial aspect of the Annals articles was that they based their recommendations on the authors’ assumptions about people’s attitudes toward meat. The idea was that, if people enjoy meat, they should not be encouraged to eat less of it, even if it causes cancer or heart disease.

While a case can be made for adjusting recommendations to avoid violating cultural taboos, attitudes toward meat-eating are not in that category and, in fact, are already changing rapidly. Despite the massive increase in popularity of meat substitutes in recent years, the authors estimated that cultural attachment to meat is too great to be influenced by healthful recommendations.

Were that the case, the authors could have concluded that there is no benefit to encouraging people to reduce meat intake. But they went further and encouraged the public to continue current unhealthful dietary practices, despite their associations with cancer and other risks.

If guidelines reinforcing meat-eating are applied also to children and to people who are unaware of meat’s risks, the opportunity to help these populations learn healthful eating habits is forfeited.

Why did the Annals decide to publish the articles?

Critics have questioned whether the Annals articles are mere clickbait, published in anticipation of a media frenzy, as happened in 2014. The journal will have to speak for itself. However, publishing tantalizing articles that lead to press controversy can influence a journal’s impact factor, which is crucial to a journal’s dominance in the competitive publishing environment, analogous to Nielsen ratings for television stations.

Conclusions

Abundant evidence has shown the health risks of red and processed meat. Despite their limitations, these new studies have indicated these same risks, suggesting that consumption of red and processed meats should be discouraged. The contrary guidelines offered by these publications are inappropriate. To the extent they are publicized, they are likely to harm public health efforts.