I am an obesity researcher and I think that in some sense the idea that we are programmed to eat and to store fat for lean days is correct, and that our current environment of abundance combines with this innate tendency to increase the prevalence of obesity. But this general statement hides many unknowns. For example, why are some people more prone to weight gain than others (in the same environment)? Why is obesity heritable (20-80% heritable depending on how you calculate it and the population you use)? Taubes is absolutely correct in stating that the "low-fat diet" mantra was promoted without any evidence to back it up and the evidence we do have seems to favor low-carb diets, at least in the short run. Fructose (and its not just high fructose corn syrup...sucrose is 50% fructose, HFCS is 55% fructose) does indeed seem to be deleterious above and beyond just adding calories, but its not the whole story either. Toxins and endocrine disruptors may play a role, but we really dont know too much about that yet. Bottom line: the notion that we know NOTHING about nutrition is false. But the notion that we know all we need to know about diets and obesity is also false. The notion that higher carb intake is responsible for most of our increased weight gain is plausible (supported by data about total carb intake in the population..its the only macronutrient that has actually increased in per capita consumption. Fat intake has declined in the last 30 years, yet weight gain has accelerated) but not the whole story either. Its also worth keeping in mind that there is no one-to-one correlation between obesity and particular disease outcomes (diabetes, heart disease). On a public health level, there is an increased risk, but there are a very large number of "healthy obese" people; there is good evidence that modestly obese adults trying to lose weight have higher mortality than those whose weight remains static or slightly increases; there is very good evidence that most diets and other prescriptions dont work and lead to yo-yo weight change that may itself be unhealthier than the baseline moderate obesity; there is good evidence that lack of fitness is a far more significant risk factor for morbidity and mortality than BMI (in adults), and so on....the "moral panic" about obesity is not always grounded in sound science. "Prevention of obesity" by focusing on weight gain in children may bypass some of these concerns, but we have to be careful to see if we are responding to a fashionable moral panic or focusing on truly evidence-based fears AND interventions (even if the fears are real, all responses are not automatically justified; some responses to the disease may be worse than the original disease, others may be ineffective and hence a waste of time and money)?

Here are the food heuristics I like to use:

• Variety, variety, variety. I don't always succeed at this when I'm not in full control of my food purchases or fall into a rut. But eating many different things is likely to capture more of the nutrients I need than eating just a few things.

• If something tastes better than it usually tastes, or seems more appealing than it usually does, eat it until it stops doing that. Flavors aren't static things that only signal what's in a food - they are also somewhat responsive to the body's needs.

• Exception to the above: If I eat a lot of something and the craving doesn't let up, it's probably not actually helping, it's just a stand-in for something nearby in foodspace that I actually need. At this time it's better to think of what it might stand in for. Really persistent legume cravings tend to go away more readily if I eat fish; if I want fruit more than usual it can mean I'm thirsty.

• Remember that food does not lose nutritional value if you add things to it. For example, broccoli is not deprived of the broccoli-nature by the addition of cheese. It's still broccoli and still contains all the goodies it contained before it was doused with dairy.

• Supplement key nutrients that my diet may not cover. For instance, I crisp up in the sun and therefore avoid it - I take vitamin D. I have a history of anemia - I take iron.

As CronoDAS has pointed out, the field of nutrition science has a long history of releasing bad studies leading to conclusions that were later proven false. Here are some of the stupid things which researchers have done, which have lead to bad dietary advice:

1. Inferring what's good or bad for humans from its effect on lab rats
2. Studying a diet's effect on one particular disease, then using it to argue for its effects on overall health
3. Conducting a controlled study, losing half the sample to dropouts and non-compliance, and pretending that the group that dropped out did so for a reason other than because the diet was hurting them
4. Measuring weight loss, but failing to distinguish between loss of fat and loss of muscle.

I believe Taubes is correct, and that the idea that low-fat diets are healthy is due to errors 3 and 4. What convinced me was hearing a bodybuilder talk about fat and muscle weight as separate things, and connecting that observation to the "rebound effect" - that is, the observation that people who lose weight on low-fat diets tend to gain it back plus extra. My interpretation of this is that some of the weight they lose is muscle, not fat, and this lowers their metabolism. This makes sense because maintaining muscle requires protein, and low-fat diets are usually also low-protein diets; the most common protein-rich foods, meat and eggs, are also high in fat.

This also explains why vegetarians don't have the same problem with low-fat diets: they're specifically instructed to be careful about their protein intake.

Great post. Since I have had (a diluted form of) the position Taubes describes for a long time (since before I was exposed to OvercomingBias, much less LessWrong) I cannot speak too much about the influences from here but it is a useful exercise to to try to trace how I update my beliefs in practice.

Caveat: I know damn well that I suck at giving true reports on what really causes me to change my mind. Our self awareness isn't particularly motivated to be honest about such things. Nevertheless I can give a best estimate on what influenced me.

If you've already thought about this, do you believe Taubes' thesis, and how did you come to this conclusion?

Yes. From what I can tell I formed the belief based on exposure to various experts that appeared to be Correct Contrarians. As someone who has taken an interest in a whole range of topics regarding health I have been exposed to experts in all sorts of fields that overlap with nutrition. It is not hard to distinguish between experts that seek out research to form accurate opinions and 'experts' who specialise in presenting authoritative beliefs. It is also not hard (given the right skillset) to independently verify the positions of such experts on their core positions. Basically none of those individuals support the idea that fat is the big culprit or a worse health risk than carbohydrate.

If this is the first time you've ever heard of Taubes, how far have you shifted your probability for the Dietary Fat Hypothesis based on reading this post?

I know little about Taubes. I've seen him as a reference but most of my exposure has been to other authors.

What more research do you intend to do to decide whether or not to continue believing it?

Probably quite a lot. ie. I am currently studying pharmacology degree with the intent of doing research in loosely related areas. I will almost certainly learn more about that kind of nutrition so as to better understand my own specific topics of interest.

How much weight do you place on the fact that I believe Taubes?

A moderate amount. Your presentation is credible and any report from an evidently educated reader that demonstrates an interest in epistemic knowledge and an understanding of bias carries weight.

On the fact that Eliezer believes Taubes?

Hmm. Eliezer does seem to be good at identifying correct contrarians so his position on such topics carries some weight. What carries even more weight is when Robin and Eliezer both support the same position - they have sufficiently different influences that agreement counts for something. I don't recall whether Robin has blogged about diet advice in particular but his writing on health topics in general give credence to this position either way.

How much did you update your beliefs based on what other commentors have said (assuming there have been any)?

To be honest probably not much. I participate in other forums with a much higher level of knowledge on this area. On health related topics where I have been disagreement with other commenters here I have updated far less than when I have disagreed on non-health topics.

One problem with all these debates about nutrition and exercise is that they mostly don't take into account individual differences. How exactly your body will react to a certain regime of diet and exercise, and which regimes are compatible with reaching and maintaining your optimal weight while feeling good and healthy -- the correct answers to these questions depend very strongly on your genotype, and possibly also on a number of entirely non-obvious and unknown environmental and lifestyle factors. (And to make things especially un-PC, the relevant genetic factors seem to correlate strongly with ethnic origin, since they've evolved significantly during the last few thousand years. This has been proven conclusively only for a few things like lactose tolerance, but it seems quite plausible that analogous group differences exist for other nutrients too.)

Therefore, except for a few simple observations like e.g. that you need some vitamin C to avoid scurvy, promoting certain diet guidelines as universally valid for all humans can't be other than nonsense. It's similar to those silly government propaganda campaigns about "safe" drinking limits that supposedly hold for everyone, whereas in reality, some people can drink a bottle of whiskey a day and nevertheless live long and productive lives (and never even appear visibly drunk), while others would be killed sooner or later by a fraction of that -- and yet others suffer from complete alcohol intolerance and get sick even from a single drink. Similarly, I have no doubt that some people exist for whom carbohydrates are as bad as Taubes says, but I also have no doubt that for many others, his claims are exaggerated to the point of nonsense, and I strongly suspect that the latter group includes a great majority of individuals in at least some ethnic groups.

As the bottom line, until scientists gain a much better understanding of the genetic and other factors involved, there really is no good way to go except self-experimentation. The only time I got into such a bad shape that I had to lose weight, I managed to devise a regime that enabled me to lose a pound a week with only a small expense of willpower, which was however completely unlike anything I've ever read from any side in the diet/exercise controversies -- and I doubt it would work for very many other people.

My conclusion has been: nobody really knows anything about nutrition, so I'm going to eat what I damn well please. (They used to say that margarine was better than butter, but now they've concluded that "trans fat" is actually worse than ordinary saturated fat.) Summing up all the various advice, it all seems to come down to "eating is bad for you." (And data on caloric restriction seems to confirm this!)

Remember this scene from Woody Allen's movie Sleeper?

Dr. Melik: This morning for breakfast he requested something called "wheat germ, organic honey and tiger's milk."
Dr. Aragon: [chuckling] Oh, yes. Those are the charmed substances that some years ago were thought to contain life-preserving properties.
Dr. Melik: You mean there was no deep fat? No steak or cream pies or... hot fudge?
Dr. Aragon: Those were thought to be unhealthy... precisely the opposite of what we now know to be true.
Dr. Melik: Incredible.

Great case study, in that studying my own reaction to your article has thought me a lot about my own decision making. And my conclusion is that reading a rationalist blog isn't sufficient to become rational!

I am thin, despite having very bad eating habits (according to conventional dietary wisdom). I had not heard of Taubes before. Specifically, I have never considered that conventional dietary wisdom could be incorrect; people say that I eat unhealthily, and I have simply taken their word for it. The fact that I continue to eat unhealthily has more to do with laziness than rationality.

How much have I shifted my beliefs now that I have read this article? I don't think in probabilities yet, but... by a lot. Probably more than I should. Neither Taubes' belief nor bentarm's belief nor Eliezer's belief nor Taubes' lack of biological credentials has affected my own degree of belief, not by much anyway. Just the fact that there exists an alternative theory of food has succeeded in demolishing the confidence I had in the conventional theory.

How could such a weak amount of evidence cause such a large shift? By relying on a bias which happened to be pushing in the same direction. Just ask me: "since you have decreased your belief in the conventional theory after reading an article on Taubes, you must have correspondingly increased your belief in Taubes' theory, right?"

Well, let me google whether the fatty food I currently eat is high-carb or low-carb, and I'll get back to you on that.

I put high levels of trust in my own repeated experience, and even moderately low fat eating leaves me feeling lousy within a day or less. This inclines me to believe that Taubes is on to something. And that people who push extreme low fat may mostly like the drama of aesceticism.

On the other hand, there are people who tolerate low fat diets much better than I do, and my general assumption is that people's dietary needs vary a fair amount.

My theory (which I don't follow consistently) is that people will do pretty well if they eat according to what will leave them feeling good three or four hours later. Some people with food sensitivities need to be thinking 2 or 3 days ahead-- sometimes it can take that long for a migraine to show up.

I realize that my theory has a bunch of caveats-- I seem to notice how I feel more than most people do. I like enough variety in my food that I have a chance to find out how different things make me feel. which, again, isn't most people's default.

I'm not going to trust anyone's theories about food unless it at least has some overlap with my experience.

This is a fascinating topic, and I hope it attracts more commentary. As Bentarm says, it is important and relevant to each of us, yet the topic is fraught with uncertainty, and it is expensive to try to reduce the uncertainty.

I do not believe Taubes. No one book can outweigh the millions of pages of scientific research which have led to the current consensus in the field. Taubes is polemical, argumentative, biased, and one-sided in his presentation. He makes no pretense of offering an objective weighing of the evidence for and against various nutritional hypotheses. He is selling a point of view, plain and simple. No doubt he felt such a forceful approach was necessary given the enormous odds he faces in trying to gain a hearing for his ideas. But the fact remains that the reader must keep in mind that he is only hearing one side of the story.

Weighed against Taubes (and others who have advocated similar positions) we must consider the entire scientific establishment, thousands of researchers who dedicate their lives to the pursuit of knowledge. To believe Taubes, we must believe that these people are basing their entire professional careers on a foundation of falsehoods. Worse, from the lack of impact Taubes' book has had on consensus opinion, we have to imagine that researchers are willfully ignoring the truths that Taubes so convincingly reveals. Nutrition researchers are intentionally lying and covering up the truth in order to protect the false dogma of the field. (Note that this is exactly the same critique of researchers made by global warming skeptics.)

I can't believe that scientists are so dishonest, or that such a cover-up could be executed successfully. I can't imagine how any young, budding nutrition researcher could go to work in a post-Taubes world with a clean conscience, if the book is really as convincing as it claims to be.

My conclusion is that to someone intimately acquainted with the field, Taubes' book is not as persuasive as it appears to the layman.

Now, I will confess that I have some independent reasons to doubt Taubes. But I would prefer not to go into that because IMO the argument I have outlined above is sufficient. Never believe a polemical, one-sided book which has been rejected by the scientific establishment. I offer that as a valid heuristic which has proven correct in the overwhelming majority of cases.

Piling on, here is Alone on diet studies:

"A glance at the methodology has more practical value than the little asterisk above a score at month 6." - regarding the uselessness and distraction of p-values.

"the purpose of these studies is not to determine the answer, the purpose of these studies is to be published"

That's actually a recurring theme in his posts. In this one he analyses the methodology of a study on a Bipolar Disorder drug. Conclusion - the authors claim the study shows that the drug is 50% more effective than a placebo. What the study actually shows, when one looks at the methodology, is that among people who responded well to the drug, 50% of them suffered relapse when the drug was discontinued. This is not surprising. And it doesn't say what the actual efficacy of the drug is at all.

Another reason diet is a good case study is that people can become very emotionally invested in their diets. Vegetarians and vegans, of course, have moral reasons, but even people with no moral reason for their abnormal diet can become incredibly defensive.

Raw foodists, paleo, low carb, high carb, organic. It's amazing to see people argue and defend incredibly poorly thought out ideas, sometimes very passionately. I'm mostly paleo myself, but I have to shake my head at a lot of the arguments and defenses I see put forth from the paleo community.

We like to say that religion and politics are mind killers. In my experience diet can be just as much so. I was actually in the process of writing a top level post about this. I'm glad this post is getting so much love though. Now I can be lazy =D

I have not read Taubes but I do read almost all the announcements on nutrition science that hit reviews like ScienceDaily etc. From what you have said of him, I probably would agree with him. I find main stream nutritional science to be tawdy if not positively dishonest. My view on fat is: trans fats are very dangerous, way too much fat is dangerous, way too little is more dangerous, cholesterol is important but dietary cholesterol is not as important (if you eat lots than the body makes less and vice versa), essential fatty acids are really essential and you will suffer without them, forget about the fat it them and like your grandmother believe that milk and eggs are great food, you need fat soluble vitamins. In general, metabolism is a very regulated feedback system. Obesity is not about the number of calories you take in. It is about how much food, which kind of food, how much exercise, what kind of exercise, gender, age, genetic makeup, poisons/disease. Carbohydrates can be made from fat and fat can be made from carbohydrate. The way to get too thin people to gain weight is not to feed them more but to alternate dieting with not dieting. I live in France where people tend to be thin. They eat very, very well and lots. They eat fatty and protein-heavy foods. They eat all the good things: garlic, red wine, fish, fruit. They eat less carbohydrate than most Western countries. I am borderline diabetic and hence my interest in this subject.

I've already thought about this, and I believe Taubes' thesis.

I came across the idea after reading about Seth Roberts experiments with eating more fats, and looked into it a bit.

I ended up changing my diet around quite a bit (almost no carbs) and have noticed quite a bit of short term improvements.

I consider myself a life extensionist and I actually found Overcoming Bias and this community via the ImmInst forums. I've read a lot about nutrition back then (2008-2009). As I can tell, so far the only guy besides me who mentioned ImmInst on LW is wedrifid. ImmInst remained my primary source of health info and I also frequently search pubmed for abstracts.

I haven't read Taubes' book. I remember faintly though that several people who disagree with the mainstream lipid hypothesis criticized the book for having very sketchy science. Its conclusions, however, seem to point in a good general direction, from what I heard. It is important to note that the "mainstream'" Taubes is against is actually very poorly established and is not something you would call solid. Why it gained momentum decades ago and haven't already collapsed in billowing smoke is a genuine mystery of our times.

I personally follow a quite carefully constructed lowish-carb diet that I tweaked over the course of the last two years, and I also take some supplements that I also researched on the net to death. I'm not really expanding my views on nutrition right now because it'd be a bit lengthy. There are several big low-hanging fruits (like vitamin d3, one that I plan to write about a top-level post here sometimes) in life extension but most of the time it is rather like a battleground of dozens of complex and ambiguous ideas.

The issue is that it requires a huge initial effort to get a rough picture about health topics, and it also requires a considerable amount of rationality to just avoid instantly going astray and settling with some snake oil, fad diet or piece of ancient wisdom. After this barrier is overcome it isn't very costly to maintain a lifestyle roughly optimized for health, according to what current science really says.

This is the key question of rationality. How can we believe what is true?

I've never read Taubes, but I share similar ideas. How did I arrive at my beliefs?

• I read a lot. I try and judge the caliber of the writer. Of course, there are idiots on both sides of every argument, but I often find that one side has a preponderance of open minded, rational, reasonable supporters. Intuitively, when the evidence isn't on your side you tend to resort to less savory tactics to win. I've found a lot of very undogmatic, well spoken people on Taube's general side of the fence.
• I find many of the paleo arguments convincing (although not as convincing as most people in the paleo community).
• I've experimented heavily with my own diet. I simply select the diet that makes me feel best. I make small tweaks constantly and big tweaks rarely (sort of a simulated annealing type algorithm). I keep the tweaks that make me feel better. My current diet would pass for paleo, although it wouldn't be typical.

Ultimately, I still don't have incredibly high confidence that I'm right. I'd love to see some reform in the nutritional science community. I'd love to see more studies. I'd love self experimentation to become the norm, so that we could all share experiences and have a reason to encourage scientists to do more diverse studies.