Last month, an op–ed in The New York Times argued that high-protein and high-fat diets are to blame for America’s ever-growing waistline and incidence of chronic disease. The author, Dean Ornish, founder of the nonprofit Preventive Medicine Research Institute, is no newcomer to these nutrition debates. For 37 years he has been touting the benefits of very low-fat, high-carbohydrate, vegetarian diets for preventing and reversing heart disease. But the research he cites to back up his op–ed claims is tenuous at best. Nutrition is complex but there is little evidence our country’s worsening metabolic ills are the fault of protein or fat. If anything, our attempts to eat less fat in recent decades have made things worse.
Ornish begins his piece with a misleading statistic. Despite being told to eat less fat, he says, Americans have been doing the opposite: They have “actually consumed 67 percent more added fat, 39 percent more sugar and 41 percent more meat in 2000 than they had in 1950 and 24.5 percent more calories than they had in 1970.” Yes, Americans have been eating more fat, sugar and meat, but we have also been eating more vegetables and fruits—because we have been eating more of everything.
What’s more relevant to the discussion is this fact: During the time in which the prevalence of obesity in the U.S. nearly tripled, the percentage of calories Americans consumed from protein and fat actually dropped whereas the percentage of calories Americans ingested from carbohydrates—one of the nutrient groups Ornish says we should eat more of—increased. Could it be that our attempts to reduce fat have in fact been part of the problem? Some scientists think so. “I believe the low-fat message promoted the obesity epidemic,” says Lyn Steffen, a nutritional epidemiologist at the University of Minnesota School of Public Health. That’s in part because when we cut out fat, we began eating foods that were worse for us.
Ornish goes to argue that protein and saturated fat increase the risk of mortality and chronic disease. As evidence for these causal claims, he cites a handful of observational studies. He should know better. These types of studies—which might report that people who eat a lot of animal protein tend to develop higher rates of disease—“only look at association, not causation,” explains Christopher Gardner, a nutrition scientist at the Stanford Prevention Research Center. They should not be used to make claims about cause and effect; doing so is considered by nutrition scientists to be “inappropriate” and “misleading.” The reason: People who eat a lot of animal protein often make other lifestyle choices that increase their disease risk, and although researchers try to make statistical adjustments to control for these “confounding variables,” as they’re called, it’s a very imperfect science. Other large observational studies have found that diets high in fat and protein are not associated with disease and may even protect against it. The point is, it’s possible to cherry-pick observational studies to support almost any nutritional argument.
Randomized controlled clinical trials, although certainly not perfect, are better tools for chipping away at causality, and they suggest that protein and fat don’t deserve to be demonized. In a 2007 clinical trial led by Gardner researchers randomly assigned 311 individuals to four groups: One group was assigned the high-fat, high-protein and low-carbohydrate Atkins diet; the second was assigned Ornish’s very low-fat vegetarian diet, which requires consuming fewer than 10 percent of calories from fat; the third was assigned the Zone diet, which aims for a 40/30/30 percent distribution of carbohydrate, protein and fat; and the fourth was assigned the high-carbohydrate, low–saturated fat LEARN (for: lifestyle, exercise, attitudes, relationships, nutrition) diet. The participants all had trouble adhering to their regimens, but all lost about the same statistically significant amounts of weight, and when compared head to head, the Atkins dieters saw greater improvements in blood pressure and HDL cholesterol than the Ornish dieters did.
The recent multicenter PREDIMED trial also supports the notion that fat can be good rather than bad. It found that individuals assigned to eat high-fat (41 percent calories from fat), Mediterranean-style diets for nearly five years were about 30 percent less likely to experience serious heart-related problems compared with individuals who were told to avoid fat. (All groups consumed about the same amount of protein.) Protein, too, doesn’t look so evil when one considers the 2010 trial published in The New England Journal of Medicine that found individuals who had recently lost weight were more likely to keep it off if they ate more protein, along with the 2005 OmniHeart trial that reported individuals who substituted either protein or monounsaturated fat for some of their carbohydrates reduced their cardiovascular risk factors compared with individuals who did not.
The other problem with Ornish’s antiprotein stance is that he lumps all animal proteins together. For instance, he wrote that animal proteins have been associated with higher disease and mortality risks in observational studies. But “Ornish is conflating hot dogs and pepperoni with fresh, unprocessed meats,” says Lydia Bazzano, professor of nutrition and epidemiology at Tulane University School of Public Health and Tropical Medicine, “and there’s a big difference between them.” A 2010 systematic review and meta-analysis of 20 studies found consumption of processed meat was associated with an increased risk of diabetes and heart disease but eating unprocessed red meat was not. A 2014 meta-analysis similarly reported much higher mortality risks associated with processed meat compared with red meat consumption and found no problems associated with white meat. The March 2014 study that Ornish cites as finding “a 75 percent increase in premature deaths from all causes and a 400 percent increase in deaths from cancer and type 2 diabetes among heavy consumers of animal protein under the age of 65,” also did not distinguish between types of animal protein. And it is worth noting that among people in the study over 65, heavy consumption of animal protein actually protected against cancer and mortality. (Also: the heavy protein consumers in the study were consuming nearly 30 percent more protein than the average American does.) “Whole foods—such as whole grain products and fruits and veggies—are healthy, but I think that dairy products, fish and lean cuts of meat or poultry can also be part of a healthy diet,” Steffen says.
So there’s little evidence to suggest that we need to avoid protein and fat. But what about the claims Ornish makes about the success of his own diet—do they hold up to scrutiny? Not exactly. His famous 1990 Lifestyle Heart trial involved a total of 48 patients with heart disease. Twenty-eight were assigned to his low-fat, plant-based diet and 20 were given usual cardiac care. After one year those following his diet were more likely to see a regression in their atherosclerosis.
But here’s the thing: The patients who followed his diet also quit smoking, started exercising and attended stress management training. The people in the control group were told to do none of these things. It’s hardly surprising that quitting smoking, exercising, reducing stress and dieting—when done together—improves heart health. But fact that the participants were making all of these lifestyle changes means that we cannot make any inferences about the effect of the diet alone.
So when Ornish wrote in his op–ed that “for reversing disease, a whole-foods, plant-based diet seems to be necessary,” he is incorrect. It’s possible that quitting smoking, exercising and stress management, without the dieting, would have had the same effect—but we don’t know; it’s also possible that his diet alone would not reverse heart disease symptoms. Again, we don’t know because his studies have not been designed in a way that can tell us anything about the effect of his diet alone. There’s also another issue to consider: Although Ornish emphasizes that his diet is low in fat and animal protein, it also eliminates refined carbohydrates. If his diet works—and again, we don’t know for sure that it does—is that because it reduces protein or fat or refined carbohydrates?
The point here is not that Ornish’s diet—a low-fat, whole food, plant-based approach—is necessarily bad. It’s almost certainly healthier than the highly processed, refined-carbohydrate-rich diet most Americans consume today. But Ornish’s arguments against protein and fat are weak, simplistic and, in a way, irrelevant. A food or nutrient can be healthy without requiring that all other foods or nutrients be unhealthy. And categorizing entire nutrient groups as “good” or “bad” is facile. “It’s hard to move the science forward when there are so many stakeholders who say ‘this is the right diet and no other one could possibly be right,’” Bazzano says. Plus, discouraging the intake of entire macronutrient groups can backfire. When people dutifully cut down on fat in the 1980s and 1990s, they replaced much of it with high-sugar and high-calorie processed foods (think: Snackwell’s). If we start fearing protein, too, what will we fill our plates with instead? History tells us it’s not going to be spinach.