A group of 14 researchers just set off a firestorm with a new series of studies that upends years of nutrition advice about meat. Their five systematic reviews, published Monday in the journal Annals of Internal Medicine, suggest there’s no health reason to eat less red meat — not even the bacon and salami we’ve been told for years to cut back on.
Led by Dalhousie University epidemiologist Bradley Johnston, the authors, who hail from seven different countries, focused on the impact of red meat consumption on cancer, cardiovascular disease, and mortality, among other effects, as well as people’s values and preferences regarding red meat.
Based on these studies, their conclusions — summarized in a new Annals clinical guideline — challenge the guidelines from just about every major national and international health group. Just four years ago, the World Health Organization’s International Agency for Research on Cancer (IARC) announced that people should cut back on processed meats if they wanted to avoid certain types of cancer. The American Heart Association and the US government’s dietary guidelines panel have also long suggested curbing our meat habit for better health.
But the authors of the new studies argue that people can “continue their current consumption of both unprocessed red meat and processed meat,” meaning whatever amount they’re currently eating. That’s because the health impact of cutting back is either nonexistent or small, and the evidence of any harms is so weak, that it’d be misleading to suggest people should avoid meat for health reasons.
Importantly, the studies did not investigate non-health reasons for eschewing beef and bacon — including animal welfare and meat production’s harmful impact on the environment — and the science backing the environmental case remains stronger than ever.
But what’s really interesting about this new series is the argument that previously published guidelines have been, well, bad science.
“These papers provide a nice counterbalance to the current norm in nutritional epidemiology where scientists with strong advocacy tend to overstate their findings and ask for major public health overhauls even though the evidence is weak,” said Stanford meta-researcher John Ioannidis, a longtime critic of nutrition science who was not involved in the research.
So it comes as no surprise that already, the Annals series has prompted a fierce blowback from various groups who’ve long argued that red and processed meat consumption should be curbed. The American Cancer Society, American Heart Association, Harvard T.H. Chan School of Public Health, and a slew of other researchers objected to the series. The Physicians Committee for Responsible Medicine — a group that’s long endorsed a plant-based diet — filed a petition with the Federal Trade Commission in response to the studies, asking the agency to “correct false statements” contained in the report, which they deemed a “major disservice to public health.”
So how did the authors of the new studies come to a wildly different conclusion? It’s less a story about whether or not one should eat meat and more about the challenges of nutrition science and how eating recommendations should be made.
In the past, many of the groups that have set guidelines for whether or not humans should cut back on meat considered a very broad range of research, from animal evidence to case-control studies, a relatively weak type of observational research. (Here’s more on different types of study designs.) As you may have guessed, there are all kinds of problems with these kinds of study designs.
Models based on animal studies don’t always bear out in humans. Case-control studies are not the most reliable, either: Researchers start with an endpoint (for example, people who already have cancer). For each person with a disease (a case), they find a match (a control) — or someone who doesn’t have the disease. They then look backward in time and try to determine if any patterns of exposure (in this case, eating meat) differed in those with cancer compared to those who don’t have cancer.
But since meat eaters differ so fundamentally from those who don’t eat meat, the reasons the two groups have varying health outcomes could have nothing to do with meat. Researchers try to control for these “confounding factors,” but they can’t capture all of them.
Some past reports on meat eating have also factored in the environmental and social effects of gobbling up steaks and bacon.
The five Annals papers did something different: They looked only at the health effects of processed and unprocessed red meat. Processed red meats — everything from hotdogs and bacon to lunch meats — are transformed by salting, curing, or fermentation. Unprocessed meats include beef, veal, pork, lamb, and venison. The papers were also systematic reviews and meta analyses, or syntheses of the research evidence that bring together a bunch of studies with the goal of coming to more fully supported conclusions. And the researchers used a very strict definition of what constituted reliable evidence for inclusion in their reviews.
More specifically, they relied on a trusted research-rating system called GRADE, or the Grading of Recommendations Assessment, Development, and Evaluation, to decide which studies to include in their papers. GRADE was developed for creating summaries of research evidence to help guide health decision-making. It’s currently the most widely used tool for evaluating the quality of science, with more than 110 organizations endorsing the method.
The idea behind GRADE is to push reviewers to base their conclusions on only the most certain evidence available. And, according to the tool’s criteria, in the case of meat consumption and health, that was large cohort studies and randomized control trials. So the researchers simply threw everything else out, including the animal studies.
The logic was simple, says study author Gordon Guyatt, a professor at McMaster University who also helped develop GRADE. “What GRADE does is say we should rely on the highest quality evidence. In this instance, we had 600 cohort studies alone.”
Cohort studies are considered to be more trustworthy than case-control studies. Unlike case-control studies, they follow people with a known exposure (eating meat) through time, waiting to see if, when, and how many people develop a particular health outcome (such as heart disease or cancer). This means researchers are not left searching for artificial controls to match their cases. And since participants are followed forward, researchers can track in real time what they’re eating instead of relying on people’s faulty memories of the past.
Randomized controlled trials, meanwhile, are deemed the gold standard in health research. They take two groups of people and randomly assign them to an intervention (in this case, eating meat or not). The idea is that the only difference between the two is the intervention (whether or not they ate meat) and not any of those other confounding factors, like socioeconomic status. And while they’re challenging (and rare) in nutrition research, they’re generally more reliable than, say, animal models.
So that’s why the conclusions of the series look different from other similar reports: They used a new approach to evaluating nutrition research, picking out the best available evidence, tossing the rest.
On a range of health outcomes — from deaths due to cancer and cardiovascular disease, type-2 diabetes, cancer incidence, stroke, all-cause mortality, and heart attack — the researchers generally found either no benefit on cutting back on meat or one so small, and based on such weak evidence, it was deemed unreliable. (You can read the papers here, here, here, and here.) For the fifth review, the researchers looked at people’s feelings about meat consumption, again focusing only on health concerns (read: not moral, ethical, or environmental reasons for avoiding meat). And they found, essentially, that many people are attached to meat, and feel being able to eat it influences their quality of life.
But the authors were clear that even the best-available evidence on meat is far from perfect. Let’s parse the language in their guideline recommendation (emphasis mine):
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).
In GRADE, there are four levels of evidence. And evidence is rated down if it’s deemed problematic for any reason — from imprecision to risk of being biased. In the case of meat and disease, the researchers determined, even the best evidence was “low certainty.”
So, Guyatt said, “We’re closer to saying: we really don’t know,” while past guidelines have generally suggested we fully understand meat’s health effects.
Now let’s look at what a “weak recommendation” means, according to GRADE. Weirdly, this isn’t just about the strength of the evidence; it’s also about people’s values and preferences.
A “strong recommendation” comes when a guideline panel believes all fully informed people would make the same choice. A “weak” one comes when “there is likely to be important variation in the decision that informed persons are likely to make,” according to a BMJ explainer on GRADE. As you’ll remember, one of their Annals reviews looked at people’s values and preferences around meat consumption, and found the majority of people value meat.
“When you trade that off with uncertain — and if it exists at all — small benefit from reducing meat,” Guyatt added, “our inference is that most people would choose to continue.” Hence, the weak recommendation.
In the past, he added, guidelines appeared to be focused on getting people to eat less meat rather than a truly dispassionate look at the science. “It doesn’t serve that goal well to point out either the uncertainty or the small effect.”
While people like the tough-to-please meta-researcher John Ioannidis called the series “very rigorous and unbiased,” others were not as impressed.
The Harvard School of Public Health — well known for trumpeting a plant-based, Mediterranean eating pattern — issued a response to the series, essentially discrediting it for discounting all the evidence showing meat’s links with poor health.
Christopher Gardner, a Stanford nutrition researcher, called the study’s GRADE approach inappropriate for nutrition. “I respect they want to have a clear-cut evidence base,” he told Vox, “but it won’t apply to lifestyle.”
Other guidelines consider observational epidemiology in additional to animal research and randomized trials, he added. “If you do that — and you’re the WHO — you say ‘based on the overall evidence from multiple disciplines, this is our best advice,’” said Gardner. “[The Annals researchers] just cut that off at the knees and said we’re not going to consider most of that.” Specifically, he was concerned that the authors threw out important and potentially relevant research, such as the PREDIMED and the Lyon Diet Heart studies. While these randomized trials didn’t focus on meat consumption, they did contain data on dietary patterns involving meat that may have been relevant.
Then there was the concern over the series’ omission: meat’s impact on climate, water, land, and pollution. “This is a missed opportunity,” the Harvard researchers wrote, “because climate change and environmental degradation have serious effects on human health, and thus is important to consider when making recommendations on diet, even if this is addressed separately from direct effects on individual health.”
But that wasn’t the purpose of the studies, said Guyatt. The point was to zero in on the fraught question of meat’s direct influence on health. Plus, he added, the new series is an attempt to do something radical: to say the rules of science should apply to nutrition. “Why have one set of rules for judging [nutrition] and another set of rules for some other area?” he asked. As he and his colleagues continue to apply their new method to other dietary questions, they may lead us to more uncomfortable conclusions.