Coffee? Keto? Olive Oil? New tool surfaces science behind diet and habits

Does eating red meat increase one’s risk of heart disease? Would eating more vegetables help? Is leaving high blood pressure untreated really a death wish? The answers might vary, depending on who a person asks, which friend or TikTok nurse, and when. Researchers at the University of Washington want to make it easier to find current, evidence-based health advice. 

A new tool from the Institute for Health Metrics and Evaluation, unveiled Monday in Nature Medicine, uses a 5-star rating system to show how much evidence exists to support some diet and lifestyle changes. The researchers analyzed hundreds of studies in hopes of helping consumers, clinicians and policymakers — awash in a landscape of wellness influencers, food lobbyists and quack advice — cut through the chatter and know the scientific consensus. The result is what they are calling the “Burden of Proof studies,” since it’s on the research to prove something is legitimate. 

Other such reviews exist, the Cochrane Library being a repository of many of them. This new tool, the authors say, is complementary to what exists, but also slightly different. Many epidemiologists assume that risk increases about the same no matter how many grams of vegetables someone eats a day, for example. “Burden of Proof allows us to understand better how the risk actually changes with consumption,” the authors said.

In medicine, “there’s always been some skepticism” about how changes to people’s behaviors can affect their long-term health, especially when it comes to recommending specific foods or activities, said Christopher Murray, senior author of the papers and founder of the IHME.

Clickbait headlines and grocery cart contents reflect the uncertainty. Cow’s milk is bad, and then it’s good. Butter — nay, all fats — must be gone, but then they’re back. Once the shopping cart is full, the Mediterranean, Keto, Paleo and South Beach diets compete for dominion on magazine covers in the checkout line. The peanut butter cups loom. (Is chocolate good or bad? Wait, what about peanut butter?)

“Diet research is really challenging,” said Jeffrey Stanaway, assistant professor of global health and lead author on the group’s analysis of vegetable health studies. It is difficult for researchers to measure how much people eat, to do so over time, and to separate their diet from other health factors (people who eat lots of fruits and vegetables are more likely to exercise, for example). 

And yet, diet and other behaviors play a significant role in disease prevention. About half of the U.S. population has a chronic condition, and long-term illnesses like heart disease, diabetes and cancer are major drivers of disability and death worldwide. “The vast majority of what makes you healthy happens outside the doctor’s office,” said Georges Benjamin, executive director of the American Public Health Association. 

By evaluating the available data for any link between vegetable eating and five different health outcomes, Stanaway could come to a conclusion: “The evidence on vegetables is pretty good,” he said. Even a conservative interpretation of the evidence, which the IHME tool uses, showed eating more vegetables is tied to a reduced risk of chronic disease, though future studies could affect that. The model is meant to be updated, and will be, as additional research becomes available, the team said. 

A three-star relationship between an increase in non-starchy, fibrous vegetable consumption and ischemic stroke was the strongest link of the bunch. Data suggest increasing vegetable consumption from one to four servings per day carried about a 23% reduction in stroke risk. The analysis also showed a two-star rating for vegetable-eating and heart disease (two on the verge of three, Stanaway said). The study did not include starchy vegetables, such as potatoes, sweet potatoes or corn, and also excluded cured and pickled vegetables (kimchee, sauerkraut). 

For the most part, dietary habits landed between one and three stars, indicating a need for more rigorous research. “I was very surprised at how many of the diet-risk relationships were much weaker” than expected, Murray said. He has a slight bit more tolerance for eating red meat after seeing those results, he said. 

All evidence on red meat and its links to disease were weak. That wasn’t unexpected to Benjamin, who wasn’t involved in the research. ​​“The things that have always been kind of fuzzy still look kind of fuzzy,” he said. 

The strongest ratings on a meat-heavy diet were two stars, for colon and rectum cancer, breast cancer, ischemic heart disease and type 2 diabetes. In the case of strokes, the researchers found a diet high in red meat could actually have some protective effects, and gave that evidence one-star ratings. Low star-ratings should be seen as areas for research investment, the IHME team said — a large, well-designed study on people with diets high in red meat could make a big impact. 

Tobacco is often the place where all of the fiery debate comes to rest. There is wide consensus among health professionals that smoking tobacco is bad for humans. IHME’s tool found evidence for strong or very strong links across eight diseases or outcomes, including larynx cancer, aortic aneurysm, peripheral arterial disease of the lower limbs, tracheal, bronchus and lung cancer, chronic obstructive pulmonary disease, and others. 

“It is irrefutable that tobacco is a major risk to health and really has a broad set of impacts across multiple cardiovascular and cancer outcomes,” all in all, Murray said. 

Still, there was less robust evidence on the connection between smoking and numerous other illnesses, including ischemic heart disease, esophageal cancer, stroke, type 2 diabetes, and others. Strangely, there was a one-star-rated link between smoking and asthma, a finding that surprised the researchers. Cannabis smoking was not included in the analysis. 

The risk of ischemic heart disease was strongly linked to high systolic blood pressure — a five-star rating — validating both common dogma among clinicians and the IHME tool’s accuracy, the researchers said in a news conference. 

The IHME team has already analyzed nearly 200 other risk-outcome combinations, ranging from alcohol drinking, air pollution and high body-mass index, to other diet factors, such as eating whole grains and legumes. Those results will be published in the future, Murray said.

Benjamin said it will take time for clinicians, policymakers and patients to see the value of this tool — the data alone might not be enough to sway the public’s understanding of risk. 

Where the rating system could be useful in the long run is the doctor’s office, when a clinician is crafting a care plan for a patient with multiple risk factors (say, smoking, high blood pressure and low vegetable consumption). If what we know about those risks can be weighed against each other, then the doctor and patient might have a better sense of what to prioritize, Benjamin said. “The less things you give people to do, the better, and the more likely they are to comply,” he said.

STAT’s coverage of chronic health issues is supported by a grant from Bloomberg Philanthropies. Our financial supporters are not involved in any decisions about our journalism.

Get your daily dose of health and medicine every weekday with STAT’s free newsletter Morning Rounds. Sign up here.