Sunday, February 3, 2013

Why Do We Eat? A Neurobiological Perspective. Part VI

In previous posts in this series, I explained that the brain (primarily the mesolimbic system) integrates various factors to decide whether or not to drive food seeking and consumption behaviors.  These include homeostatic factors such as hunger, and non-homeostatic factors such as palatability and the social environment.

In this post, I'll examine the reward system more closely.  This is the system that governs the motivation for food, and behavioral reinforcement (a form of learning).  It does this by receiving information from other parts of the brain that it uses to determine if it's appropriate to drive (motivate) food seeking behavior.  I covered its role in motivation in the first post of the series, so in this post I'll address reinforcement.

Behavioral Reinforcement

In the last post, I discussed palatability, or the pleasure associated with eating food, and how that influences food intake.  But I saved a key element of the discussion for this post: behavioral reinforcement.  Behavioral reinforcement is a form of learning that happens when you engage in a behavior that the reward system decides is favorable.  The reward system initiates a program that makes you feel good, and increases the probability that you'll engage in that behavior again.  Conversely, if you engage in a behavior that the reward system considers unfavorable, you may develop an aversion to it and your likelihood of engaging in that behavior will decrease.  This is a key part of our species' ability to survive and navigate a constantly changing world.  It allows us to adapt to novel environments rather than rely exclusively on hard-wired behaviors.

The reward system evolved to reinforce behaviors that enhance survival and reproduction, and oppose behaviors that reduce survival and reproduction.  But it's important to remember that the reward system didn't evolve to handle a world of ice cream sundaes and pornography.  It reinforces behaviors that would have increased survival in the ancestral environment*.  That's why it favors calorie-dense easily digested food, and tends to oppose unnecessary physical activity.  In the ancestral environment, the risk of suffering reproductive consequences due to excessive consumption of energy-dense food and too little physical activity was probably low-- the main concern was eating enough to sustain your calorie expenditure.  Here's Dr. Bruce King** from a very recent article (1):
The gastrointestinal, sensory (taste and olfaction), and brain feeding mechanisms that developed during the past 2 million years were highly adaptive for ancestral hunter-gatherers living in an environment with limited high-density foods and periods of food deprivation. Today, however, humans in industrialized countries live in what has been called an “obesogenic environment.” The non-homeostatic brain reward circuitry that was acquired during evolution to seek out and eat as many nutritionally high-dense foods as possible is able to overrule the physiological inhibitory mechanisms that were designed to limit meal size and weight gain.
Regardless of the evolutionary justification, we know today that certain food properties reinforce behavior in animals and humans.  If a food contains these properties, in general when you eat it repeatedly the reward system will make you more likely to obtain and eat it over time, sometimes even if you aren't hungry.  The reward system makes the food more palatable and more tempting, reinforcing consumption behavior.  Here is a list of some of the core factors that drive reward.  You may recognize them from the post on palatability:
  • Sweetness
  • Fat
  • Starch
  • Calorie density
  • Salt
  • Free glutamate ('umami')
  • Absence of bitterness
Let's illustrate this with some research, then an everyday example.  Dietary fat and calorie density are two of the most rewarding food properties, and they are linked to one another since fat is the most calorie-dense macronutrient.  In an experiment published in Physiology and Behavior in 1991, children were given yogurts that contained novel flavors (i.e. it was the first time they had encountered those flavors), in either high-fat or low-fat form (2).  Researchers found that at the end of the experiment, children ended up preferring flavors that were previously paired with the high-fat yogurt, even if they were no longer associated with high fat.  The reward system caused them to 'learn' to prefer flavors associated with fat and calorie density, which subsequently drove their eating behavior.  Similar findings have been reported using beverages fortified with different amounts of carbohydrate (3).

Conversely, if a food is paired with nausea, the reward system will learn to avoid the food and cause it to seem less appealing (4, 5).  This is called a learned taste aversion.


Now, a common sense example.  Initially, people tend to prefer sweet milk chocolate because sweetness is an inherently preferred food quality (from birth), and the bitterness of dark chocolate is inherently disliked.  As they're exposed to chocolate repeatedly, dark chocolate often becomes more pleasant as the brain learns that dark chocolate contains more drug than milk chocolate, and is not toxic despite its bitterness.  This is the same process that occurs for coffee and beer, both of which are bitter but contain a habit-forming drug.  The unpleasant aspects are overcome as they're associated with positive properties over time, and behaviors to seek and consume those foods/beverages are reinforced.  The same thing happens for many foods, for example, we acquire a taste for vegetables over time if they're repeatedly paired with fats and other calorie sources.  No one would ever acquire a taste for kale if we ate it plain every time-- it's bitter and contains virtually no calories.

Chocolate is an archetypal high-reward food, and most people reading this would probably feel tempted if a square were right in front of them, regardless of hunger (including me).  That's because it's a killer combination from a food reward perspective: extremely high energy density, high fat, sweetness, a nice texture, and a drug.  It doesn't get much more rewarding than that.  But the world isn't black-and-white: dark chocolate also seems to have meaningful health benefits if eaten in moderation.  Still, it can facilitate overeating and fat gain.  My solution is to keep a bag of plain roasted cocoa nibs in my kitchen.  They taste good, but not good enough that I'll eat them if I'm not hungry.

This brings us to another important point.  I've noticed that my writing on the subject of food reward has frequently been misunderstood and/or misrepresented, e.g. "we should eat food that tastes bad", or "we should reduce food reward and palatability as much as possible".  I think a lot of that is coming from people who don't want to understand food reward.  But for those who do, I'll clarify.  Food reward and palatability are factors that influence food intake and body fatness.  They aren't inherently bad or good-- they're tools that you can use to achieve your goals, whether that's fat loss, weight gain, or eating food that you enjoy.  There are other ways to lose or gain fat, and food reward doesn't erase those-- it only adds to them.  Food reward is one of the reasons why many of us eat beyond our true calorie needs, but that doesn't mean it needs to be ruthlessly minimized in all situations.  Life is a balancing act, and enjoying food is part of the equation.

The Updated Model

We haven't uncovered the last 'module' yet, so the model looks the same as last time:

* Whatever that is exactly.  But I think we can agree it's not 2012 USA.

** Bruce King is a seasoned obesity researcher currently at Clemson.  He wrote an outstanding review paper titled "The rise, fall, and resurrection of the ventromedial hypothalamus in the regulation of feeding behavior and body weight" that I often refer to (6).  The VMH is one of the key hypothalamic nuclei that regulates food intake and body fatness.  I highly recommend this article to anyone who has a serious interest in the VMH, and the hypothalamic control of food intake and body fatness in general.


pawpaw said...

Recently heard a story on NPR? about the rise of caffeine in foods, such as Cracker Jacks. (Caffeine levels need not be disclosed in the labeling.) Research with children showed they developed a preference for the caffeinated versions of these processed foods. Another method to steal market share from your competition?

Stephan wrote: "No one would ever acquire a taste for kale if we ate it plain every time-- it's bitter and contains virtually no calories." While I accept this point as true for many, may I plead for a more level playing field. I'm a market gardener, growing and selling greens this time of year such as kale. After multiple frosts, the sugars in these and other greens can be quite high (as cellular antifreeze), bugs are gone so the bitter compounds are quite low, and our market customers quickly buy out such high quality greens once we give out samples and they're habituated. Kale and chard flew off the tables at our market today, most of these folks return customers. Some anticipating the joy of sharing such high quality greens with Christmas guests, working to limit that holiday gain to less than 6 lbs. Family (ancestral?) traditions inform these choices as well. And we have guest chefs and other programs at our market to introduce kids to veggies at their peak.
Typical conversation with new greens customer: What do you add to these greens when cooking them? Grower: nothing, our greens hold their own, try them. Customer: but which oil do you cook them in? Grower: None. Customer: Garlic or onions to make them more palatable? Grower: None! Customer: You must be adding something? Grower: Steam! Try it!

From 2 grower families with 10 children between us: we each cook large (16 qt) pots of greens, nearly always eating them all. One of my daughters rejoices when she gets seconds and thirds of these greens.

My point: Quality vegetables, grown well in their preferred season and consumed fresh, can hold their own against many (though not all) processed foods. Especially when the processed foods are limited by various strategies. The fresh ginger snaps came out tonight only after the soups, whole milk and greens were gone, and we limited the cookies to one each.

OK, off my soapbox now (over)defending kale grown well. We do imbibe bitter chocolate, though more rarely than our greens.
And we spoke about food reward at dinner tonight; thanks for teaching us through this series.

Gretchen said...

Stephan, Thanks. Nice to have overviews presented clearly.

Re kale: I agree it can taste good undoctored if fresh (I grow it too, to eat, not to sell), but few Americans have access to fresh kale, and most people prefer greens with added fat.

For that reason, the low-fat fad may have caused people to eat less vegetable foods.

I *like* bitter foods like black coffee, bitter greens, and bitter melon. As babies, I'm told we all liked coffee (teaspoonsful, not huge mugs). I assume one can acquire a taste for a whole class of tastes.

Hepoberman said...

Is there indication of nutrient/mineral "deficiency solved" reward mechanism(s)? I wonder if a re-prioritization occurs under deficiencies. For example, would a sodium deficient animal still prefer high calories over high salt when deficient in salt ? It seems to me, a system capable of re-aligning preferences based on need is a very complex system indeed. Can these "need" signals be distorted or going 'unheard' due to insulin/leptin resistance ?

Great article, Thanks!

Kris said...

Great article, Stephan... I've been following this series closely.

It's funny that I was just holding a cup of coffee when I read the part about coffee being bitter but due to the drug-like effect the brain would learn to enjoy it over time.

When I first started drinking coffee, I just had to use both milk and sugar.. quite a lot of it actually. Black coffee tasted downright nasty.

Today, all I drink is black coffee. I've actually learned to love the taste of it. Very interesting how my preference has changed based on the rewarding effects of the coffee.

Gys de Jongh said...

We might be looking at the wrong end of the problem :)

Diabetologia. 2012 Sep;55(9)
Rectal taurocholate increases L cell and insulin secretion, and decreases blood glucose and food intake in obese type 2 diabetic volunteers.
PMID: 22696033

Chris Wilson said...

Hi pawpaw, you bring up an interesting point that I'd like to discuss with Stephan sometime. We know pretty well that in herbivores at least, post-ingestive feedback and reward can drive beneficial nutrient and medicine seeking behavior, not only for plant primary metabolites but for secondary metabolites as well as minerals.

I think we're only just beginning to understand how these systems might work. It reminds me of the Adelle Davis nutrition experiments where toddlers, when presented a menu of strictly real foods, seemed to select foods based on fluctuating physiological requirements which is perfectly consistent with the literature from herbivores.

I think the key is to connect and educate this instinctive/reward system in the context of an environment with at least a minimal diversity of strictly real foods. Obviously, the whole processed food environment plays havoc with this.


pawpaw said...

See this link (or search "Farm Fresh Kids") for a program that empowers kids to choose their own whole foods each week, all summer long. Kids often sample before buying:
We joke about kids selecting veggies, fruits and recipes their parents aren't used to, but are glad to see it. Veggie chips are a hit for some averse families, with their crunch, oil, and salt, but still a whole food.

Alexander Rinehart, DC CCN said...

I wonder if you've covered food accessibility and cost and whether you include this in food environment considerations. I could see a low-cost, highly available food (like a candy bar in urban convenience store), as a factor in food reward.

ex. Seeing ice cream go "on sale" at the grocery store and feeling compelled and rewarded when you buy it.

Anonymous said...

A very interesting post

Unknown said...

hey is that " if we dont eat we will not die".......ive heard this some where
<a href=">

Brandon Belcher said...

I found out a shortcut to Lose 5 lbs in 5 days and its EASY! So try it!

Unknown said...

Fantastic article, it’s so helpful to me, and your blog is very good,
and you put it in a way that everyone can understand.
check out this review.


vdiggity said...

Stephan, great series. I've been a long time reader for the past 5 years, but am now a first time commenter!

Curious to hear your take on the impact of gut bacteria and fungi (the "organ" in our body comprising our skin, mouth and gut biomes) on the neurobiology of our eating.

Others (Chris Kresser, Art Ayers, Emily Deans, et al) have long touted the importance of healthy gut bacteria and the resulting impact on mental and physical health, including but not limited to obesity. Additional research has long shown a definitive close connection between the gut and the brain (with complex feedback loops).

In your framework, how does the gut biome impact our neurobiology through the vagus nerve, the hypothalamus, or other means?

Are we looking at a future where fecal transplants are the most effective medical interventions for obesity and diabetes? Hopefully there are less drastic options, such as ingestion of properly fermented foods (which could be highly palatable - kefir, or not - kimchi/kraut).

Below are examples of reading I mention above:

Reijo said...

Very nice and pragmatic recap on reinforcement. This series is valuable resource to myself as a dietitian.

I have seen data that claims sugar being the most rewarding, not fat. Perhaps there is some variation between individuals.

'The new finding is that, after control for the presence of other macronutrients, the relative reinforcing value of food is only related to sugar intake'

Unknown said...

Nice post on Diet topic as we eat whenever we feel hunger or appetite for food.

<a href="”>Easy Diet Plans</a>

kulimai said...

"... I'll clarify. Food reward and palatability are factors that influence food intake and body fatness. They aren't inherently bad or good-- they're tools that you can use to achieve your goals, whether that's fat loss, weight gain, or eating food that you enjoy." I see. Now I know what reward means and what palatability means.

Much interesting and often thought provoking info on this blog, but Stephan, your consistent refusal to even try to provide a defensible definition of your crucial terms continues to make it all just a kind of storytelling. There might well be more in this, but we'll never know if you don't make a serious attempt at the necessary conceptual clarification.

Jane said...

kulimai, I can't believe you said that. Stephan is constantly defining his terms. He must be sick and tired of doing it. I'm actually astonished he still does this blog in spite of all the flak he gets.

His problem is that he is writing both for scientists and for non-scientists. The scientists will get him if he makes statements that go further than the data, and the non-scientists will get him if he doesn't. He has to choose his words very carefully. Considering that, his blog is a masterpiece of clarity. But then, I've been reading it every day for years, and I know what he means in a way a casual reader might not.

kulimai said...

Hi Jane, thanks.

"food reward to refer specifically to the motivational value of food"

"Food reward is the process by which eating specific foods reinforces behaviors that favor the acquisition and consumption of the food in question"

"food reward is measured by the ability of food or food-related stimuli to reinforce or motivate behavior"

so food reward is a value or a process or an ability -(presumably it should not matter which) that is a property of a given food? Is it reasonable to assume that food really has a reward property (value/process/etc) that is constant across consumers/ environments/ food contexts/ times?

"It is not a tautology or circular reasoning to say that the reinforcing value ...of food influences food intake" In that case what is a tautology?

Perhaps I am missing something important and enlightening, in which case I would appreciate if someone would direct me to it.

Unknown said...

Happy New Year! Hugs! Good Post...

Jane said...

Hi kulimai
Yes, there is certainly a problem here. But the problem actually consists in the assumption that Stephan thinks food reward is THE explanation for obesity, which he does not. Hope I'm not putting words into his mouth here. As I understand it, he would like obesity to be due to nutritional deficiencies but thinks there is no concrete evidence for this. I believe there is, but I can see how this evidence would appear equivocal to him and to other scientists.

We shall have to wait and see. Papers are being published all the time which shed light on this subject. You can be sure that Stephan is reading them.

kulimai said...

Jane thanks again, I appreciate your input. But either there is some hidden clarification we did not notice or there is something deeply flawed here. It does not matter if food reward is *a* cause or *the* cause of something, in either case for the statement to be evaluable as true or false we need to know what food reward is. Hunt as much as I may, I cannot find reasonable and useful non-circular definition on Stephan's blog.

Suppose I made the claim that the beauty of an object is a factor that enters into its collectibility potential and proceeded to say that I understand (define) beauty as a value of an object that contributes to making people collect that object.

Have I actually made an empirical claim ie. a statement of (putative) fact? It is easy to see that I have not, there is no instance of collecting or not collecting an object that could show that I am wrong. I am just using words, without making a statement that could be true or false. Like I submit that a black raven is black. It does not make me scientist.

To have a real claim/theory about food reward Stephan must give it a *reasonable* definition that is *independent* of the claim he is making about food reward. I am not saying this is impossible, but obviously it is hard, and Stephan has not done the work but seems to fudge by alternately using two ill-defined notions of palatability and reward. Sorry by the way that I have to be negative, --I like Stephan too.

Chris Wilson said...

kulimai, I think you are deeply confused. Read over Stephan's many posts and dig into the subject a little more. The reward system is in the brain (and is studied pretty extensively). Food interacts with it via a variety of its properties/qualities. There is no need to search around for a Platonic definition of Food Reward.
There is a growing empiric study of how food interacts with the reward systems, and there is plenty of latitude for individual variability.


Josh said...

The idea that food reward & palatability influence obesity is not Stephan's theory, it is an entire field of scientific research. It is not up to him to define terms, these have already been defined and agreed upon by other scientists. In his blog he summarises some the research that has been done in this field and explains some of the implications in a way that is easy for the lay person to understand, and he does a very good job of this. He has explained and re-explained the definitions of food reward and palatability so many times it is ridiculous considering that it is such an easy and intuitive concept to understand.

Sanjeev said...

> or there is something deeply flawed here

The circularity allegation has been refuted many times.

Even if part of the theory was circular, the fact that it's been triangulated by researchers and studies from several different fields - opiate blockade studies, neurotransmitter studies, many rat intervention studies, fMRI studies completely vitiates the allegation - it's just nonsensical at this point

Sanjeev said...

> He has explained and re-explained the definitions of food reward and palatability so many times it is ridiculous

jeez ... no kidding; how hard is it to type this in google? circular reward

which gives you this: how hard?

Sanjeev said...

Another interesting search string: circular reward kulimai

inreresting, first hit more than 16 months ago ...

here's the above link to copy and paste to guarantee work safety:,or.r_gc.r_pw.r_qf.&bvm=bv.1355534169,d.aWM&fp=ade955b4265491c9&bpcl=40096503&biw=1436&bih=876

Gretchen said...

As Stephan has noted, "One of the problems with food reward is it's defined differently by different people, even researchers, and often it isn't defined at all."

He noted that a better term would be "action-selection system."

I think of the reward *system* as being a little computer in the brain that receives input from many sources, neural, chemical, and psychological, integrates this input, and decides on the best action at the current moment.

So, I might love cheesecake, and cheesecake is available to me right now. But if I'm about to be attacked by a rabid wolverine, I probably won't sit down and eat that cheesecake. Note I said "probably."

But some people use the term "rewarding" as synonymous with "reinforcing," and some use it synonymously with "palatable."

The field would be less confusing if someone would come up with better terms.

kulimai said...

Welcome, defenders of the faith. Yes, I am deeply confused and will remain so, until someone tells me in non-immediately circular and reasonable terms what food reward is. In principle of course you could do it in terms of all the research you cite. But you don't even try.

Indeed it has been claimed many times that the circularity allegation has been refuted. A single convincing refutation would suffice. I cannot find it.

A proper and explicit understanding (as opposed to a strong but vague belief of having that understanding) of what a claim is being made about is a necessary condition for making an empirical claim. I am genuinely sorry if this comes through as unreasonable or offensive.

Chris Wilson said...

Kulimai, you wrote:
"A proper and explicit understanding (as opposed to a strong but vague belief of having that understanding) of what a claim is being made about is a necessary condition for making an empirical claim. I am genuinely sorry if this comes through as unreasonable or offensive."

My friend, you are on a long and hopeless quest for a Platonic Form. Let it go. The circularity argument has been refuted, for all intents and purposes. Moreover, there is no problem of definition. I am genuinely confused why you think there is.

The claim is simple: the hedonic and reward value of food influences brain regulation of intake and body fat-ness. This can be defined and studied in terms of reward pathways in the brain, much as research in drug addiction has advanced.

You seem to think that "Food Reward" is a property inhering in the food itself- it is not, it is the effect in the brain's reward pathways arising from eating a food or foods. I fail to see why this is problematic from a definitional point of view.

kulimai said...

Chris, (my friend), I am most impressed by expressions like Platonic Form, although I admit I do not quite see how they would help to answer the question I am asking.

Also I have no idea why you think that I think "that "Food Reward" is a property inhering in the food itself", -- I don't. I was quoting Stephan.

Now here is what you offer:
"the hedonic and reward value of food influences brain regulation of intake and body fat-ness. This can be defined and studied in terms of reward pathways in the brain, much as research in drug addiction has advanced"

Since I asked about reward value I assume you mean here that at least reward value can be defined in terms of brain pathways. Quite possibly true. Once this is done (but not before) we can check if any claim we make about it is true or false. So could you please just provide this definition or point me to where Stephan provides this or endorses a definition of this sort? Thanks.

Jane said...

I didn't explain myself properly. Food reward is neither 'a' cause of obesity nor 'the' cause of obesity. It's just a shorthand term people use to refer to a certain body of research.

If I understand you correctly, you think it must be defined properly before conclusions can be drawn about its role in causing obesity. But nobody is saying it causes obesity. I expect you have encountered people on the web claiming Stephan thinks food reward causes obesity. They have a lot of fun at his expense, but in the end the laugh is on them.

kulimai said...

Jane, yes, a shorthand to evoke associations is fine for a thought provoking conversation, but it does not make a theory.

Really I am just looking for an empirically testable claim about food reward, which I submit with some confidence, cannot be made without an explicit and nonambiguous understanding of what food reward is. That is all.

Stephan writes many edifying, interesting, useful and enlightening things. We can be all be grateful. No question about that. He may have theories of various things too. But as things stand, he does not have a theory about food reward. He might remedy this by saying what food reward means or saying that his theory is actually about something else. Many vociferously obfuscate the issue by claiming without demonstrating that it does not exist. I rather doubt that they are really helping him, though that may well be their intention.

Chris Wilson said...

kulimai, in response to a quote you attribute to Stephan you wrote,
"so food reward is a value or a process or an ability -(presumably it should not matter which) that is a property of a given food? Is it reasonable to assume that food really has a reward property (value/process/etc) that is constant across consumers/ environments/ food contexts/ times?"

This is the confusion I was clarifying by telling you that there is no putative Food Reward inhering in food- rather it is studied as an interaction between foods and the brain's reward system. This also necessarily involves psychological concepts like "motivation". It's a complicated process, to be sure, which is why we'll never be able to study and measure it with, say, "units of food reward", if that's what you're after.

I'm sorry if I've offended you, but, again, I'm still not sure what the problem is. I think you're insisting on a philosophical problem existing that isn't there.

I better leave off this discussion now...


RLL said...

Actually there are negative food reactions. And I think that they are widely known if not studies. People mention food aversions after eating things immediately before some sort of severe gastro reactions. I have read of these being associated with food poisoning and even chemo-therapy.

Stephan Guyenet said...

Hi Kulimai,

I have addressed this numerous times, quite clearly, so how about we reverse the question. Please explain how one would argue that food reward IS a circular concept. I have never understood the rational basis for this argument.

kulimai said...

Hi Stephan

I never actually said (nor do I believe) that the concept of food reward cannot be defined, neither did I say that it is necessarily a circular concept. I said that (although you have indeed addressed the issue many times) the definition you gave was circular. This is a quite different claim.

I do not claim or imply to have a theory of food reward, or to have made empirical claims about it. If I did, however, then indeed I would have to give a non-circular, reasonable and explicit characterization of what food reward is. (And emphatically, it would not be sufficient for me to claim that such a definition is in principle possible in place of providing the actual definition.) Otherwise quite inevitably my claims and theories about food reward would remain empirically untestable. I think this point is crucial. Without a proper understanding and appreciation of it much unnecessary confusion is generated and furthermore, and even more importantly, there is little hope of finding the appropriate remedy.

Jane said...

Why should it be necessary to have a theory of food reward to find the appropriate remedy for obesity? What's wrong with facts?

kulimai said...

"Why should it be necessary to have a theory of food reward to find the appropriate remedy for obesity?" Well, often a theory of what goes on helps to find the remedy. But I fully agree, it is not at all always necessary. Often we find a remedy and we have no idea how it works.

But I have never said a theory is always necessary. What I said was that it is wrong is to think we have a claim or theory when we do not. This typically impedes any genuine progress and can easily cause harm.

Claims about food reward that sound like empirical, factual claims are not, unless we are told explicitely what food reward is. If I put forward the hypothesis that ravlens are black you can check if I am right by looking at ravlens. But if I don't tell you what ravlens are or just say ravlens are black things, or say I have defined them many times and I won't bother again and anyway they clearly *can* be defined in principle --should I hope that you will ever take me and my theory of ravlens seriously?

"What's wrong with facts?" Nothing, at least in the sense you are asking. What is wrong is to believe that a statement is about facts when it really says nothing about them.

I repeat, I also think Stephan is onto something important. But factual-sounding statements that are not will not help him or his audience.

kulimai said...

Jane, I have an eerie feeling you understood my earlier sentence
"there is little hope of finding the appropriate remedy" as 'there is little hope of finding the appropriate remedy for obesity'. No, I did not mean that, I meant, 'little hope to remedy our thinking, little hope of finding the true factual claims to substitute for the one that is not'

Stephan Guyenet said...

Hi Kulimai,

OK, well then let's rephrase the question: how is my definition of food reward circular?

Jin said...

Hi Stephan,

Thank you very much for your blog. I've found the information here useful and quite frankly, life changing.

I've learned so much about Real (traditional) Food, and learning about Food Reward has opened my eyes to the current (obesogenic) Food Environment we live in.

I've changed my eating habits, and I have renewed respect for the brain's Reward System and all its urgings.

Thank you, and wishing you good health in the new year.

kulimai said...

Hi Stephan

It would be helpful if you provided a definition that you are happy with, otherwise I might be shooting at shadows. But look at this for example:

(1) "food reward is measured by the ability of food or food-related stimuli to reinforce or motivate behavior"

You don't say what behavior, but clearly we don't care here much about grooming, exploring etc. so one must assume this refers to food intake behavior. Ok, so we measure food intake behavior, or equivalently on this level of discussion we measure food intake. That will give us at least a relative measure of food reward, at least in the context of a given experiment. (I don't quite see how this helps in understanding what food reward means in general, across experiments but let us put this aside.)

So 'Food reward' or 'the reinforcing value of food' is measured in terms of food intake, it is a food intake value.

(2) "It is not a tautology or circular reasoning to say that the reinforcing value ...of food influences food intake" So you are saying in effect that 'it is not a tautology or circular reasoning to say that the food intake (value) influences food intake (value)'.

I do think minimally this needs to be clarified.

Stephan Guyenet said...

Hi Kulimai,

OK, thanks for clarifying. Your objection is essentially the same as the others I've discussed this with.

The key point is that the reward value of a food is not measured by how much of that food a person eats. If that were the case, the argument would indeed be circular.

Food reward is measured by the motivational value of food, or how hard an animal or person is willing to work to obtain it. In rats for example, you can use a progressive ratio schedule, which basically measures how many times a rat is willing to press a lever for a drug or food. A similar test can be applied in humans. Often researchers use a measure called the relative reinforcing value of food (RRVfood). It measures how hard people are willing to 'work' for food relative to non-food items. There are computer 'games' that researchers use to assess this, rather than making people push levers. Importantly, food reward is never measured by the quantity of food people eat.

Food reward is also measured by the reinforcing value of food, or to what extent an animal or human 'learns' to prefer it over time. A rat is given a food (or drug) in a particular location. Researchers then measure how much time the rat spends hanging out in the same location in the future, hoping for another piece of food or drug. This is called 'conditioned place preference'. In humans, conditioned preferences can be measured by looking at how people 'learn' to prefer certain novel flavors over others when they are paired with rewarding nutrients like fat and starch. Importantly, when these conditioned preferences are assessed, this is done in a context where they are no longer paired with the rewarding nutrient (or are equally paired across conditions).

Does that clarify?

kulimai said...

Stephan, a very happy new year to you.

Thanks for taking the time to explain this. Indeed on this version we have no circularity, but at the price, it seems to me of running quickly into a self-contradiction.

(1) For reward to be relevant to obesity, it is clearly not enough to say that we eat more of rewarding food, we also need to assume something to the effect that 'more calorie-dense/ fattening /palatable food is more rewarding', i.e. on your definition we are willing to make a greater effort to obtain it.

(2) Now consider yourself or me or indeed anyone who follows your advice to eat less rewarding food. All of us, by hypothesis, would work less for more rewarding food, we would make a smaller effort to obtain it, than less rewarding food, since that is precisely what we decided to do. Do we then work less for food for which we work more?

(3) Are you going to say that (a) it is impossible to follow your advice since doing so would contradict the food reward hypothesis of obesity that you want to maintain or more reasonably (b) it is possible to follow your advice but the food reward hypothesis of obesity is not generally correct?

My point is that adopting a behaviorist definition of food reward seems to add a layer of unnecessary serious conceptual difficulties.

(N) One might of course be able to devise experiments where people under the effect of drugs, distraction, whatever are behaving like mice, not being able to follow your advice. But that would only give a food reward theory for mice-like humans, not for people.

Jane said...

Yes, I was putting words into your mouth. Sorry.

Jin said...


To review, are you aware of any system in the brain that detects and regulates nutrient content & needs?

On the surface, it seems to me we would not have such a system otherwise I'd be craving beef liver & bone broth, not chocolate chip cookie and red wine.

TexasBruin said...

Hi Stephan,
Today (Jan 2, 2013) I read an Associated Press story regarding recent research re: brain activity associated with fructose. In summary, MRI brain imaging showed that fructose intake did not trigger the response that glucose did. According to Yale endocrinologist, Dr. Robert Sherwin, glucose "turns off or suppresses the activity in the areas of the brain that are critical for reward and desire for food." Fructose does not have this effect. Earlier this year you blogged about fructose, but with this recent finding have you reconsidered your position on fructose?

I sincerely enjoy your blog.

Gretchen said...
This comment has been removed by the author.
Gretchen said...

We've been having bad weather, and I haven't wanted to drive to a grocery store. I have plenty of meat in the freezer, but I'm running low on vegetables.

So I decided to harvest the last stalk of brussels sprouts. It had fallen down, and we had a foot of snow, and I couldn't remember exactly where it was, so I was digging in the snow to find it when it struck me: I'm like those rats that do uncomfortable tasks in order to get rewarding food.

Not sure I would have run across an electrified barrier, though. Brussels sprouts aren't *that* rewarding.

Laura said...

Thank you Stephan! This is the best explanation of food reward yet. I have been reading your blog for quite some time and not "getting" it... I hope you have a great year!