Does fat matter?


As of September 2017, new Sweat Science columns are being published at Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


Just a decade ago, our concept of “healthy eating” was so simple and straightforward: fat is bad. These days, not so much. Amby Burfoot’s most recent Peak Performance blog post summarizes the key points from “The Great Fat Debate” held among four highly respected nutrition experts (Walter Willett, Alice Lichtenstein, Lewis Kuller, and Darius Mozaffarian) in the current issue of the Journal of the American Dietetic Association.

There was plenty of disagreement, but some common ground. For example, total fat is less important than the type of fat: saturated fats (e.g. dairy and meat) are less desirable than unsaturated fats (e.g. olive oil). But replacing fats with processed carbs isn’t the answer, and will probably make things worse — which brings up the fundamental problem with this kind of debate. As Harvard’s Mozaffarian puts it:

Dietary recommendations that focus on selected nutrients, such as total fat or saturated fat, are often confusing for the public, result in illogical dietary decisions, and increase the potential for manipulation of nutrient targets by the food industry… If we’re eating an otherwise healthful diet including plenty of vegetable oils, fruits, vegetables, fish, and nuts, it will be much less important what the saturated fat level is.

Or as Lichtenstein (of Tufts) puts it, more simply:

I think we have to stop talking about individual dietary components because when one goes up another goes down.

Given the continuing disagreement about fundamental questions (is cholesterol bad?), it seems pretty clear to me that we don’t have enough understanding of the complex relationship between diet and health to successfully micromanage the ratios of specific nutrients. On the other hand, we have pretty unambiguous evidence about the benefits of certain patterns of eating — like getting lots of vegetables and fruit. Until the research is a little less murky, that’s the approach I’m sticking with.

4 Replies to “Does fat matter?”

  1. As you may know, Stephan Guyenet pins the blame for obesity on highly rewarding foods, reckoning that industrially processed foods are “professionally designed” to be highly rewarding. In this view, whether dietary fat is fattening would depend on the context in which it is prepared, and not just its inherent nutritional characteristics.

    His most recent post indicates a striking correlation between the rise in obesity and the rise in the proportion of fast food in the diet (which he is taking to be a proxy for industrially prepared food):

  2. Interesting stuff — thanks for the link, Phil. I definitely agree that food engineered to be highly rewarding is likely a key element of what obesity researchers refer to as the “obesogenic environment.” No doubt whatsoever — and a very good reason not to get bogged down in micromanaging ratios of fat, carbs, good fats, bad fats and so on.

    That being said, I’m suffering from a bit of “correlative fatigue” after viewing endless graphs showing that Factor X rose sharply sometime after the Second World War and thus caused our current obesity crisis. It seems to me that the vast majority of social, demographic, technological and environmental factors have changed dramatically in the past century, and even more dramatically in the last half-century. But I don’t think that Moore’s Law, for example, is making us fat.

    Actually, let me rephrase that. Moore’s Law is making us fat, to the extent that technological changes have enabled us to become dramatically less active (and enabled better control of information flow to make fast food cheaper, for that matter!). It’s one among many, many factors that have changed, none of which, on their own, hold the key to our problems.

  3. When you can fatten rats more effectively by feeding them more powerful processors, then Moore’s law will be pari passu with reward as an “obesogenic” candidate. Psychologically, a plenitude of similar things (“correlations”) weakens the interest any any particular example. But logically, that is a fallacy: you cannot weaken the evidence in favour of a cause by finding other correlations that are self-evidently non-causal or reverse-causal (pants size?) A million such examples would not suffice.

  4. But is Moore’s Law really “self-evidently non-causal”? The technology made possible by mushrooming processing power has turned our society into a sedentary service economy. The near-complete absence of physical effort in the average daily life is a perfectly plausible physiological mechanism for societal weight gain! 🙂

Comments are closed.