Hacker News Reader

" there is a not a single study in the history of science which has been able to demonstrate eating at a caloric deficit and gaining weight."

That's absolutely right.

"takes what is a relatively simple issue and makes it needlessly complex."

Human metabolism and psychology is complex. Basically, it's the most complex system that we know of.

You can't just throw out psychology and metabolism, and say it doesn't matter, because as long as you eat less calories than you consume you're alright.

People have know "calories in vs. calories out" for decades. People are still getting fatter. Maybe if you shout it a bit louder, it will start working? Sorry about the sarcasm, but that's how I feel about the issue.

I feel it's a bit like telling a depressed person to just "stop it". Snap out of it! Why are you thinking these bad thoughts? It's bad for you! Stop it!

https://www.youtube.com/watch?v=Ow0lr63y4Mw

You do touch on this though "(how to ensure someone sticks with the program)", but then seem to gloss over it. You start out saying it's "not complex". If it's not complicated, why is it not solved yet?

What you eat, is certainly not everything, but sugar drives behaviour very strongly. More so than other nutrients is my impression. It affects motivation which again affects how well you stick with a program. It's not everything, but it's part of the puzzle. Other things that affect compliance could be, mental health, friends, advertisement, culture, economical situation.. How is this not complicated?

I think we totally agree. My entire point is that this is enormously complicated, but mostly in the psychology department, and not so much in the eating department. Eating a caloric defect works for everyone, so the difficult part is figuring out how to get them to do that, which seems to be a psychological issue entirely.

Again, we agree that it is enormously complex, I am lamenting the fact that we don't put more resources into the psychological aspect of it.

Of course eating at a calorie deficit will work for everyone if your only goal is 'lose weight'. If you actually want to be healthy it's more complicated than that.

Maybe, but there is a massive amount of low hanging fruit in terms of making people healthy by simply making them less fat.

Have you considered that what you eat might make a difference to your psychology?

Of course, what I'd like is more studying of that and less studying of whether or not different macronutrient compositions cause you to lose more weight through something other than calorie deficit.

I think the point is that the macronutrient compositions is the driver in how food affects your behavior.

That is, what you are asking for is a study that would show why large populations of people can't just "eat less." Why are they stuck in a rut of eating more than they need? Worse, why are many of them not even happy about it, but still can't get out?

This hypothesis is basically that that behavior is fueled by a feedback cycle that has, in large part, the actual foods that people eat caught in it somewhere.

Now, I think I see what you are getting at. Could you fix this at other parts of the feedback cycle? Perhaps. And I certainly agree that should get research, too. I just don't think this research is unrelated.

> I think the point is that the macronutrient compositions is the driver in how food affects your behavior.

That is absolutely not the point of Taubes's research. Reread the article.

Sorry, my point is more that the study would be basically the same anyway.

That is, if this singles out some food as more likely to cause overeating/fat gain, that will still be an incomplete explanation. Which is where the psychology aspect would step up.

Right?

> Sorry, my point is more that the study would be basically the same anyway.

I disagree, the study would be completely different. This study is studying the acute effects of diet on the body, specifically trying to determine if macronutrient composition has an effect on weight loss, given equal calories.

The study I'm talking about would deal with long term behavioral aspects of dieting, and would look for people who had long term success with dieting and see what they are eating. Or maybe there would be some other study design, but it certainly wouldn't be like the one they are doing now, which completely controls diet thus eliminating the compliance aspect.

The reason I think this distinction is so important is because of the amount of energy being consumed by the scientific community around debating the question of whether or not macronutrient composition has a direct effect on weight loss. Even if the people who feel like macronutrient composition has a direct impact on weight loss are right, the effect is not very dramatic.

It would be much more fruitful to study the psychological impact of specific diets, and whether or not they are likely to impact over eating, which is the one and only cause of obesity, as you mentioned.

You don't think they'll be keeping behavioral notes on the subjects? They are keeping them under observation for the full study, right? Seems logical to record behavior, as well.

Though, I do have to cede the point that a proper study of this sort would be different.

"Show me an obese person not eating significantly too much."

http://en.wikipedia.org/wiki/Cohen_syndrome http://en.wikipedia.org/wiki/Bardet-Biedl_syndrome http://en.wikipedia.org/wiki/Ayazi_syndrome

There. That was easy.

I don't think genetic disorders should be considered as they're exceptions to the rules.

I stand by the statement "Show me an obese person (with no genetic conditions) that eats less energy than they use".

It doesn't exist, because it's not physically possible

I just showed you specific genetic conditions that lead to excess body fat without over-eating. You have to reckon that. Science doesn't work that way. You can't just ignore evidence that counters your hypothesis, waving your hands and saying "pah, these exceptions to the rules." On a similar note, do you think most females develop breasts because they consume more calories than they expend (v.s. males)? By your line of reasoning, you should believe that.

"Like many other medical conditions, obesity is the result of an interplay between genetic and environmental factors. Polymorphisms in various genes controlling appetite and metabolism predispose to obesity when sufficient food energy is present."

http://en.wikipedia.org/wiki/Genetics_of_obesity

Over-eating is required.

None of your links actually support the idea that you can become obese without eating excess calories. They are all genetic diseases where obesity is a symptom of the disease. Maybe because they eat too much?

When talking about "the obesity problem" in America these people are a statistically insignificant population. Less than .1% of people have those 3 diseases combined, whereas something like 30% of American adults are obese.

> This seems intellectually interesting to me but it's frustrating because I feel it takes what is a relatively simple issue and makes it needlessly complex. Yes, maybe there is some amount of optimization you can do with your diet, but the simple fact is that there is a not a single study in the history of science which has been able to demonstrate eating at a caloric deficit and gaining weight. Show me an obese person not eating significantly too much.

Taubes would say that an obese person eats too much precisely because he is fat, not the reverse. The arrow of causality has been flipped. His body is demanding increased caloric intake (via insulin, leptin, and ghrelin) to maintain fat stores. Those fat stores can't be liberated when insulin levels remain high. If this hypothetical obese person keeps eating carbs and has become insulin resistant, those fat stores aren't going anywhere without a fight. A calorie restricted diet will help some, as less food produces less insulin, but it's far easier to restrict insulinogenic foods. This allows the fat stores to be utilized, and weight loss is the natural result.

It sounds counterintuitive at first, but this theory states that the human body knows how to maintain caloric intake. That's what hunger is for. But it can only do this if we allow the endocrine system to function normally.

Before dismissing the idea, listen to Taubes's talk: http://youtu.be/ywRV3GH5io0

Taubes would say that an obese person eats too much precisely because he is fat, not the reverse. The arrow of causality has been flipped. His body is demanding increased caloric intake (via insulin, leptin, and ghrelin) to maintain fat stores.

That "arrow" of causality is more of a messy tangle. Everything (appetite, eating habits, body fat, environment, mental state, exercise/lifestyle habits) affects pretty much everything else.

In particular the environment can have a rather strong effect on eating habits. So be unreasonable, and adapt the world to support your goals.

But if a fat person simply eats less, no matter the macronutrient content, he or she will lose weight. This is well documented, and even Taubes doesn't disagree with that. There may be ways to lose weight more quickly, but my point is that the challenge here is how to make people who are on terrible diets consistently stick to any decent diet with a caloric defecit, not the optimal diet. If someone can't comply with a low carb diet it quite frankly doesn't matter if it's more optimal that a high carb (though I disagree with Taubes generally and think his analysis misses the forest for the trees).

> But if a fat person simply eats less, no matter the macronutrient content, he or she will lose weight. This is well documented, and even Taubes doesn't disagree with that.

A low-carb diet doesn't have to be intentionally low calorie. That's the point. Eating very low carb allows the body to adjust hunger levels to match existing fat stores. If you're fat and you allow your body to do its job, you won't be as hungry because you have plenty of energy stored up and the body is eager to use it. In other words, we've evolved to survive times of plenty and times of famine, and our modern diet with extremely high carb intake throws this system out of whack.

Low calorie is just forced starvation and although you will always lose weight this way, it's an uphill battle because you're fighting the body's fat regulation mechanism. You might be eating high-carb but low calorie, causing insulin spikes, extreme hunger, and protein wasting. It's not an easy diet to follow and it might even be counterproductive.

This is the theory anyway. This queston has been posted to Taubes many times over though, and he himself can explain it a lot better than I can. If you're interested in the topic, this is definitely worth a read: http://garytaubes.com/2012/11/what-would-happen-if-thoughts-...

> Low calorie is just forced starvation and although you will always lose weight this way, it's an uphill battle because you're fighting the body's fat regulation mechanism. You might be eating high-carb but low calorie, causing insulin spikes, extreme hunger, and protein wasting. It's not an easy diet to follow and it might even be counterproductive.

This is based on a false dichotomy of dieting sustainably low carb or unsustainably starving yourself.

Let's say we have someone who is 6' and 260lbs. To get to that state you need to consistently eat a diet of around 4000 calories. If that person eats 2000 calories of ANYTHING for the rest of their life, which should be completely sustainable, they will not only lose the weight (down to about 180 lbs), they will keep it off.

On the other hand, if that person continues to eat 4000 calories but they change their diet entirely to a ketogenic diet, guess what: they will still be fat.

The point is that when we are dealing with the 30% of Americans who are now obese, the low hanging fruit is in simply eating less. There may or may not be a suboptimization in macronutrient composition, I'm not saying that's definitely wrong. I'm just saying there's more low hanging fruit than that.

Most obese people have not been obese their entire lives, so obesity can't be a root cause of their obesity. At most, Taubes' writings might help explain why some obese people have trouble managing their over-eating.

> From the practical standpoint of actually trying to lose weight/get people to lose weight, the challenges in nutrition are almost entirely around compliance (how to ensure someone sticks with the program) rather than substance (what people put in their bodies).

Compliance and substance, using your definitions, are not orthogonal. "How you ensure someone sticks with the program" is not independent of what the things they eat are.

I get it why techies -- myself included -- would like the world to have nice, neat, orthogonal levers where you push this input and you get this simple output, and you don't have to manipulate a bunch of inputs to get a simple output (and, don't get a bunch of complex changes in output when you change a simple input.)

Unfortunately, many interesting domains -- especially those involving the human body and particularly the interaction between body and behavior -- don't work that way, and preferring that it works more simply isn't going to change that.

> if we spent more time studying how to teach compliance I think we'd be making a lot more progress towards stopping obesity.

The thing is, there's considerable evidence that "compliance" problems are neither one-size-fits-all nor simply a matter of teaching. Much current effort is on compliance, but it often involves how to manipulate metabolic signals to help people be better able to comply (there's definitely education as a component of that, but its not alone.)

And we understand where there are positive feedback loops involved, and breaking those is part of that (that's a major reason for the various forms and relatives of gastric bypass surgery -- it enables radically breaking some of the feedback loops that make it difficult to acheive compliance.)

> "How you ensure someone sticks with the program" is not independent of what the things they eat are.

We have not put nearly enough scientific resources into figuring that out to say that so confidently.

> Much current effort is on compliance, but it often involves how to manipulate metabolic signals to help people be better able to comply (there's definitely education as a component of that, but its not alone.)

I don't agree that much of the effort is on compliance, there is a massive imbalance towards substance from what I can tell, but who knows maybe you're right.

The argument is that "compliance" is significantly influenced by hormonal regulation, which is influenced by amount of carbohydrates consumed.

If it were that easy, we wouldn't have an obesity epidemic. I get that the solutions for individuals seem straightforward, but something is affecting most of our population, and it's problematic.

Let's say that they're not be looking for "the perfect personal weight-loss diet," to sell, but working toward larger systemic changes that will reverse or halt the trend in the greater population (like mandating vaccines, but I doubt the answer here is a "fat vaccine").

Taubes touches on it in this article a bit: http://garytaubes.com/2011/11/catching-up-on-lost-time-%E2%8...

It's not a scientific study, but it addresses the obese people not eating significantly too much.