On March 12, 2012, for instance, there was a profusion of especially scary headlines, including one in the New York Times: “Risks: More Red Meat, More Mortality.” This story referred to a research finding that just three additional ounces of red meat a day were associated with a 12 percent greater risk of dying overall, including a 16 percent greater risk of cardiovascular death and a 10 percent greater risk of cancer death. The study’s announcement echoed around the world, with news reports in virtually every country.

The data for that report came from the so-called Nurses’ Health Study II, which has followed more than 116,000 nurses for more than twenty years and is among the longest and largest epidemiological studies ever undertaken. For the red meat analysis, researchers at the Harvard School of Public Health, which directs the study, combined the nurses’ data with a similar, smaller data set on male physicians from another epidemiological study they oversee. In the questionnaires answered by these doctors and nurses, the investigators discovered an association between eating red meat and reduced mortality. However, an association, as we know, can be merely coincidental—it does not demonstrate cause and effect, and this association, it turned out, was tiny.

The actual numbers underlying the 12 percent finding (percentages often look more dramatic when they are calculated from small numbers) show that the increase in the risk of dying was only one person per hundred over the twenty-one years of the study. Moreover, the risk did not rise in lockstep with meat eating (meaning that eating a certain amount more red meat didn’t translate smoothly to a certain amount of increased risk, which is that “dose-response” relationship that epidemiologists consider crucial for establishing the reliability of an association). Indeed, the risk associated with red meat eating in the Harvard study dropped steadily as meat consumption grew, and then only worsened in the group of biggest meat eaters—an odd finding that suggested there might be no real association after all.

But what about that group of biggest meat eaters? Could they not be seen as a cautionary tale? Many other observational studies have shown an association between eating a great deal of red meat and negative health outcomes. Possibly a high consumption of red meat triggers an effect only seen at a very high threshold? Or, more likely, maybe this effect is seen because people consuming a lot of red meat today are living less healthy lifestyles overall for reasons that have nothing to do with meat. In choosing to eat a lot of red meat, most of these people have consistently ignored the linchpin of dietary advice from doctors, nurses, and health officials for decades. It’s quite likely, therefore, that these people are failing to prioritize their health in other ways: they probably don’t visit their doctors regularly, don’t take medications, don’t exercise frequently, attend cultural events, or embed themselves in meaningful ways in their communities—all factors that have been shown to be associated with good health. It is therefore not surprising that in the Harvard study, the top meat eaters were also found to be less physically active, more obese, and more likely to smoke.

By the same token, it is also true that people eating a lot of fruits and vegetables over the past few decades are healthier in ways that have nothing to do with diet. People who make a conscientious effort to follow doctor’s orders, whether to take a pill or exercise more regularly, have long been found by researchers to be healthier than people who don’t. This effect, called the “compliance” or “adherer” effect, was discovered during the Coronary Drug Project in the 1970s, when researchers found that the men who took the intervention drug most faithfully cut their heart disease risk by half. But surprisingly, men taking the placebo most faithfully also cut their risk by half. The objective value of the intervention mattered less than the willingness to follow the doctor’s orders. It turns out that people who dutifully follow advice are somehow quite different from the sort of people who don’t; maybe they take better care of themselves in general. Maybe they’re richer. But whatever the reason, statisticians generally agree that this compliance effect is quite large.

Therefore, any associations found between meat eating and disease, in order to be meaningful, must be big enough to overcome this compliance effect as well as other confounding variables. Yet, like the small association that Harvard researchers found in their 2012 study, the associations seen between red meat consumption and heart disease have generally been minimal, a scientific detail that study leaders tend not to emphasize and that the mainstream media have also, on the whole, overlooked.