Defending the truth about diets

In late February, Dr. Salim Yusuf, a prominent cardiologist, gave a talk at the Zurich Heart House, circulated on YouTube before it was pulled. Citing observational data of his own,Yusuf asserted, after noting that he has no expertise in nutrition, that in effect, everything we know about diet and cardiovascular disease is wrong.

He told the world that eating fish is neutral, eating vegetables useless, and that heart disease rates go down as meat-eating goes up. If only he had recommended smoking cigarettes, the bizarre, surreality of it would have been complete enough to rival that famous scene in Sleeper.

Of course, this is wrong — and strangely, for a researcher who has written many times on matters of epidemiologic methodology, wrong at the level of Epi 101.Yusuf was citing observational data across many countries. The sine qua non of interpreting such data reasonably is the avoidance of the ecological fallacy, in which A and B both happen but are “true, true, and unrelated.”

For instance, there is much less Ebola where there is more 4G LTE cell phone service, and more Ebola where cell phone service is poor, patchy, or absent. This, of course, is notbecause 4G LTE cell towers protect against the Ebola virus, but because Ebola is endemic in poor, rural areas of sub-Saharan Africa that lack such towers. There are innumerable examples of similarly nonsensical associations.

Sticking with only the most obvious of rebuttals: meat is a very small part of the diets in poor countries. As countries become more affluent, they can afford, and — sadly for the health of people and planet alike — generally choose to eat more meat. But affluence doesn’t just procure meat. It also procures medicine, and technology. Countries that eat more meat have more cardiologists to prescribe drugs, and perform angioplasties; they have more cardiothoracic surgeons doing coronary bypass procedures.

One obvious way to confirm or refute the merit of cross-sectional data is to look for change over time in those very behaviors in a given population, and then ask if the results were as predicted.

Cultural transitions in both India and China have shifted traditional, plant-based diets to more meat-centric, Western type diets, and rates of chronic disease in general, diabetes and heart disease in particular, have skyrocketed. These massive examples, alarmingly on display for all the world to see, belie Dr. Yusuf’s conclusions.

So does a study, just published in JAMA, which looked at dietary components associated with mortality in the U.S.

Controlling carefully for other relevant variables, this study reaffirms the associations long backed by the weight of scientific evidence and sense alike: increased risk with highly processed foods, and reduced risk with vegetables, fruit, nuts, seeds, and fish.

We actually know some things about diet and health. These days, that view is often attacked as defense of dogma. But truth is not dogma; it’s just true. It’s not dogma to defend heliocentrism against excessive scriptural literalism; it is irrefutable fact. And though many fewer accept it as such, so is evolution by natural selection. So is climate change. Scientists throughout history have at times died defending truth objectionable in their day, and the forces that burned them at the stake were the forces of dogma.

I presume we can agree that when zoos give meat to the lions and eucalyptus leaves to the koalas, it is not in the service of dogma.

Rather, it actually is possible to know some things based on consistent observation and sensible interpretation in context. We know what to feed dogs and dolphins, porcupines and penguins and tropical fish accordingly. We have the same bounty of observational information about Homo sapiens.

We have data from randomized trials, too, showing again, and again, and again much the same thing. Across populations and places around the world, decades and generations and research methods; diets rich in vegetables, fruits, whole grains, beans, lentils, nuts, and seeds and plain water for thirst lower rates of chronic disease and premature death, and diets lower in these and higher in meat, soda, and processed foods do the opposite.

The defense of this contention, which enjoys the support of a global who’s who in public health, owes nothing whatsoever to dogma — but rather is richly informed by data. It is a defense against dogma, a defense of both science and sense where they converge in a veritable sea of supporting sources.

Those of us defending what we truly do know about diet and health generally need not fear being burned at the stake these days for our pains. But since diet and lifestyle are the leading determinants of years added to or taken from lives, and life added to or taken from years — there certainly are lives at stake in this debate. For that very reason, science and sense must prevail.