Friday, November 29, 2013

Food Reward Friday

This week's lucky "winner"... Oreo cookies!!!



It's been hard to find time for Food Reward Friday lately, but this week I'm bringing it back with a vengeance.  Oreos have been one of America's favorite cookies since their introduction in 1912 by Nabisco.  They boast a tempting combination of crunchy chocolate cookie and soft, sweet filling.  From a food reward perspective, the combination of concentrated sugar, starch and fat make them a winner.  From a nutritional perspective, these same factors make them a loser.

Although most adults can think of other desserts we'd rather eat, I think many of us can remember a time when Oreos were just about the pinnacle of deliciousness.  More than tasty enough to throw a tantrum for.  It also helps that Oreos lend themselves to playing with your food.

Interestingly, Oreos were the subject of a recent study on food reward in rats (1) that WHS reader Chris Cappuccio passed on to me.  This study, presented at the 2013 Neuroscience conference, got a lot of media attention, both positive and negative.  What did it find?

Food reward is the process by which behaviors are reinforced by food, and a "rewarding" food is one that makes you more motivated to obtain it once you've eaten it a few times.  In the current study, researchers used a measure of reward called conditioned place preference.  Basically, they repeatedly gave rats Oreos in a particular location, and then measured how much time the rats subsequently spent hanging around in that location hoping for Oreos*, even when the Oreos were omitted.  Crucially, the Oreos were pitted against healthy bland rodent chow, not against an absence of food.

They found exactly what has been shown many times before: rats LOVE human junk food.  The rats trained with Oreos preferred to hang around the Oreo-associated location even when no Oreos were there, demonstrating that the Oreos are highly rewarding (i.e. Oreo consumption increased the likelihood of Oreo-seeking behavior).  Oreos are also highly rewarding in humans, otherwise they would not have sustained human Oreo-purchasing behaviors over the last century.

What comes next is the experiment that made this study so controversial: they compared the reward value of Oreos to the reward value of cocaine and heroin in their experimental model.  Intravenous cocaine or heroin was paired with a specific location, pitted against intravenous saline at another location, and the researchers subsequently measured how much time the rats spent hanging around the side paired with drug even when no drug was administered.

As expected, rats like cocaine and heroin, and they spent more time hanging around the drug-associated location hoping for a hit.  Here's the critical finding: in this experiment, the reward value of heroin and cocaine was not significantly different from the reward value of Oreos.  Also, all three conditions caused similar effects on neuron activity in the key reward-related brain region the nucleus accumbens.  The authors end the abstract with this:
These findings suggest that high fat/sugar foods and drugs of abuse trigger brain addictive processes to the same degree and lend support to the hypothesis that maladaptive eating behaviors contributing to obesity can be compared to drug addiction
This is what triggered the avalanche of positive and negative media coverage.  The negative coverage was mostly from people who are tired of hearing the argument that reward and addiction are the same thing, or that activation of brain regions involved in reward means something is addictive.  These people have a point, but there's a baby in the bathwater that we shouldn't discard.

Drug addiction is basically a very strong motivation to obtain the drug, as a result of the drug's high reward value (due to direct actions on brain reward circuits).  A person is so motivated to obtain a drug that she has a hard time controlling drug-seeking and drug use behaviors, and these behaviors supersede constructive behaviors like maintaining relationships and being financially responsible.

Similarly, people can be highly motivated to obtain certain foods (ice cream, pizza, potato chips, Oreos) due to the high reward value of those foods.  Although calling this 'addiction' is controversial, it does lead to the consumption of unhealthy calories in excess, superseding constructive behaviors like eating an appropriate quantity of healthy food.

The word 'addiction' triggers strong emotional reactions and quickly leads to unproductive debate in the popular media.  I think the best way to think about this study is to discard the word 'addiction' and simply think about reward.  Oreos, cocaine, and heroin all had a powerful ability to motivate behavior in rats.  Even though we can't do this study in humans, and I doubt humans would find Oreos as rewarding as heroin, it is nevertheless clear that Oreos and other similar foods are highly rewarding in humans and can lead to behaviors that undermine health and other positive life outcomes.  This study is a good reminder of that.


* In a manner of speaking.  We can't read their minds to see if it truly contains hope.  See comments below by Aaron Blaisdell.

18 comments:

Ronald Pottol said...

Imagine what it would have been like with the old trans fat Oreos! I know back when I ate such things I found the new ones to be no where near as good.

Chuck Currie said...

Well said...should be required reading for those on both sides of the argument.

Garrett MacDonald said...

“Drug addiction is basically a very strong motivation to obtain the drug, as a result of the drug's high reward value (due to direct actions on brain reward circuits).”

So the addictiveness of a substance is driven by its reward value?

“A person is so motivated to obtain a drug that she has a hard time controlling drug-seeking and drug use behaviors, and these behaviors supersede constructive behaviors like maintaining relationships and being financially responsible. Similarly, people can be highly motivated to obtain certain foods (ice cream, pizza, potato chips, Oreos) due to the high reward value of those foods. Although calling this 'addiction' is controversial, it does lead to the consumption of unhealthy calories in excess, superseding constructive behaviors like eating an appropriate quantity of healthy food.”

Your definition of addiction requires a substance to have high reward value, so if calling a drive to consume foods with high reward value “addiction” is controversial, then people who don’t believe that it constitutes addiction must be operating with a different definition of addiction from your own.

“The word 'addiction' triggers strong emotional reactions and quickly leads to unproductive debate in the popular media. I think the best way to think about this study is to discard the word 'addiction' and simply think about reward. Oreos, cocaine, and heroin all had a powerful ability to motivate behavior in rats.”.

This feels like sleight of hand. You can’t say drugs and junk food are similar by calling them both addictive without backlash, so you say drugs and junk food are similar by calling them both rewarding.

Ann Anagnost said...

I would love to know the original recipe compared to what they are today. I remember the Oreos of my childhood as more delicious than today, but maybe that is nostalgia? But I can also imagine a chain of substitutions as the food chemistry industry became more sophisticated.

R said...

Nicely put. I fall on both sides of this argument. Food reward is clearly powerful and worth our attention, but anyone who has watched a loved one's life be utterly ruined by hard drugs, or watched the process of withdrawal from said drugs is going to take issue with the term 'food addiction'. A distinction needs to be maintained.

Kris said...

Great article as usual, Stephan.

As a recovering drug addict and now a recovering food addict as well, I can tell you that the two are exactly the same. The social consequences aren't as severe and you can still function in life as a food addict, but the thought processes and symptoms are the same.

I don't think Oreos are as addictive as Cocaine or Heroin though (not to humans anyway), that is kind of ridiculous although it makes for a great headline.

But junk foods in general are addictive enough to cause real problems for a lot of people. Junk food addiction can ruin people's lives, just like alcohol or drug addiction.

I personally think the words "addictive" and "addiction" should be used - because that's exactly what this is and these are words that most people understand.

Of course, not everyone becomes addicted though, these foods can lead to changes in behavior without leading to full-blown addiction. But the same could be said for alcohol and no one really doubts that alcohol is addictive.

Grinch said...

Stephan, if the brain responds to food and drugs similarly, why would it be controversial to consider both addictive? That just doesn't make sense to me. Is the idea that gambling is addictive also controversial?

Back to Oreos, one idea I have is that Oreos at least in my case help discredit the carbohydrate insulin hypothesis. Proponents will quickly say that Oreos are fattening because of their sugar or carbohydrate content, but then you have people like me who are perfectly capable of gaining fat by overeating high carb foods (pizza, ice cream, pasta w/alfredo), but at the same time I don't really like Oreos (or any cookies really) so I will seldom eat them, certainly not in excess. Same goes with soda, I simply do not like sugary soda. If sugar and/or carbs per se had some hormonal effect promoting over-consumption and increased fat stores, I would think all processed junk should affect me the same way. Its only specific foods that are palatable to me, and usually contain both fat and carbs that give me trouble.

Galina L. said...

For a person not used to that level of sweetness from a childhood and the chocolate/mint combination Oreos are not rewarding at all. May be rats just react on the high amount of fat , starch and sugar together ignoring the taste? I had to discreetly spit one Oreo in a paper napkin during a party because I could not finish it. I didn't follow a LC diet them ant ate sweet things from time to time, but Oreo was way too sweet.

For a human degree of sweetness and saltiness matters a lot, and in order to be tasty, food should hit a particular taste spot. I looks like it partially depends on a habit. Steffanson reported Inuits didn't use salt, and members of his expedition also un-learned to desire salt in their food in a while. Many people who follow a low-carbohydrate diet start to notice a natural sweetness of raw almonds and green beans they didn't detect before, and pears and grapes became too sweet.

Stephan Guyenet said...

Thanks for the interesting comments everyone.

Hi Garrett,

You asked "so the addictiveness of a substance is driven by its reward value?" Yes, almost by definition, because reward is what motivates, and addiction is a heightened motivational state. If a person weren't highly motivated to seek and use the drug, he wouldn't be classified as addicted. All known addictive drugs act directly on reward circuits in the brain, most notably dopamine signaling. Food, sex, and other "natural" rewards act on the same circuits, but they do so indirectly via sensory input rather than by direct chemical actions in the brain.

Hi Ann,

Nice to hear from you. I don't enjoy today's Oreos very much either, but when I was a kid they were amazing! My guess is that my own tastes have changed rather than the flavor of Oreos. I'm sure there have been recipe changes, but they probably did extensive testing to ensure that the flavor is the same.

Hi R,

Good point.

Hi Kris,

Good thoughts. One of the big objections to the term 'food addiction' is that we have to eat food to survive. We have no physiological need for heroin or gambling, so out-of-control behavior related to these is easy to classify as addiction. Also, I think it's just hard to compare a person who loses control around food sometimes to a heroin addict-- the two are on a different scale.

Yet although we need food to survive, we don't need pizza and ice cream to survive. Most people don't lose control around plain oatmeal or raw almonds; they lose control around highly rewarding/palatable foods.

That's why I prefer to think about food reward rather than food addiction. Whether or not they're literally addictive, we know that certain foods can motivate unhealthy behaviors.

Hi Grinch,

It's a good question. It's controversial precisely because there are two sides to the issue. Some people would agree with you. In fact, I think most would agree that food CAN be addictive in some people, but that most overweight/obese people don't qualify as food addicts.

What makes this question so difficult to answer is that the definition of addiction is arbitrary. If you fall on your wrist, your bones are either broken or they aren't. But there is no clear physiological definition of addiction. Addiction is defined by a cluster of behaviors that were chosen to identify the point at which the habit-forming effects of a drug have become severe enough to warrant treatment. We know it when we see it, at least in severe cases, but it's quite squirrely to pin down.

If you use a very liberal definition of food addiction, e.g. a person consumes certain foods beyond hunger and this has negative consequences on that person's life, then most people would be classified as addicted. If you use a stringent definition, e.g. requiring among other things that a person exhibit severe withdrawal symptoms when they don't get their favorite food, then very few people would be classified as food addicted.

Often in medicine you have to create arbitrary cutoffs for the purposes of treatment (e.g. LDL cholesterol over 130 mg/dL-- even though there is no meaningful difference between 129 and 131 mg/dL), and that's what happened with addiction, but the concept has limitations.

Patrick S said...

A great article, fantastic to read and very informative. I even feel worse now when eating oreos or similar candy. ;P
But there are times, when i just need to have some chocolate. I know the reason now, too. ;D
It is just so hard to force u to not go to the fridge then.
Do you guys maybe have some tricks when the need for sweet(s) comes?

Greetings

aaron blaisdell said...

Hi Stephan,

Thanks for using the "Food Reward Friday" column to educate us about how to interpret experimental studies. I spoke with the PI at his poster at SfN, and he was very shy at first, wary of receiving more flack for his study. When he realized I was genuinely interested in his research, he opened up to be a very nice and thoughtful individual. He was apologetic about how out-of-proportion his findings were portrayed by the media. The only reason the media discovered his study despite not being published yet is because it was reported in his university news paper as an example of undergraduate research being conducted there. He had no intention to go to the media with his research in such a premature stage, let alone to have it spun in into the claim that "oreos are just as addictive as cocaine and heroin."

One correction to your post, however. You state "and then measured how much time the rats subsequently spent hanging around in that location hoping for Oreos". I agree with everything in that statement except the "hoping for Oreos" part. That implies an S-O or R-O association. While the rats might have formed such associations, performance on the conditioned place preference task does not provide sufficient evidence to make that claim. It may be merely the S-R association (Thorndike's Law of Effect) that motivate the rats to hang around on the Oreo-paired side of the arena.

Stephan Guyenet said...

Hi Aaron,

Thanks for the comment. I can imagine the media frenzy must have been quite embarrassing for him.

Regarding your comment on the "hoping for Oreos" statement, you've exceeded the limits of my psych knowledge. I don't know what S-O and R-O associations are. Of course I'm stretching a little when I say they're "hoping" for Oreos, since I can't read their minds to see whether or not they feel hope. This was basically my attempt to put it into human terms and make it more understandable.

Anyway, I'll put a note in the post directing people to your comment. If you have a chance to expand on your comment further it would be useful for me to read.

Matthew Green said...

Any black friday stories? :)

aaron blaisdell said...

Sure thing. In the field of instrumental learning, a form of associative learning, S stands for Stimulus, R for the instrumental Response, and O for Outcome (the reinforcer). Back in 1898, Thorndike conducted the first experiments investigating instrumental learning in animals (and later applied it to people, especially in the field of educational psychology). His research led him to the conclusion that all that is needed for an animal subject to learn a new instrumental response, such as a rat pressing a lever followed by food reward, was for the formation of an S-R association, that is, an association between the conditioning box (S) and the lever press response (R). Thus, by receiving a food reinforcer each time it pressed the lever, an S-R association was strengthened. The strength of the S-R association in turn determines the likelihood that the rat will press the lever (R) in the presence of the stimulus (S), which is the conditioning box in this example. He claimed that while the outcome (food reinforcer) was necessary to "stamp in" the S-R association, the outcome did not actually enter into the association. That is, it is as if the rat was pressing the lever whenever it was in the box, but was NOT EXPLICITLY pressing the lever TO GET FOOD. It was pressing the lever merely because of the strong association between being in the box and pressing the lever.

--end part 1---

aaron blaisdell said...

-start part 2---

Half a century later, some theorists questioned the assumption that the outcome does not enter into the association during instrumental learning. Some clever new procedures were developed that showed evidence that the subject, in some cases, actually DID act is if the outcome was part of the associative representation motivating behavior. For example, if a rat were trained to press the lever with a food reinforcer, and then the food was devalued (either by satiating the rat by giving it free access to consume the food right before a test session, or by conditioning a taste aversion to the food, such as with lithium chloride injections into the stomach which induces gastric malaise), the rat would no longer press the lever. This suggested that once the outcome acquired a new (negative) value, it no longer would motivate the behavior. The researchers who discoverd this effect suggested the rat did form an S-O and/or an R-O association. That is, an association between the box and the outcome (S-O) or between pressing the lever and getting the food (R-O). Sometimes, devaluing has no effect on an established instrumental behavior. One example is if the lever pressing is overtrained, say be allowing the rat to press the lever for food over 20 daily sessions, then devaluing the food does not diminish lever pressing. In these cases, lever pressing is said to be purely driven by habit learning (the Thorndikian S-R association), and not by the R-O association. Many of our behaviors that are very well trained can become habitual, which is why devaluing the outcome or other attempts to reduce our motivation for a substance can fail to reduce our behavior to acquire that outcome (e.g., drugs, food, etc.).

Kent Berridge at the University of Michigan, and Bernard Balleine who used to be in my department at UCLA but who is now at the University of Sydney, have pioneered the behavioral neuroscience of the habit and goal-directed instrumental learning systems, as well as the role of Pavlovian S-O associations in motivating instrumental behaviors (e.g., relapse, etc.). Thus, in my circle of behavioral psychology and learning theory, we are very cautious about the language used to describe motivated behavior. We resist the temptation to attribute behavior to goal-driven mechanisms (e.g., S-O and R-O associations, like pressing the lever TO GET FOOD) unless it is empirically warranted. Otherwise, we must remain agnostic with respect to the habitual versus goal-directed nature of the behavior.

Of course I understand that you were using the goal-directed language as an expedient to convey your meaning to a general audience, and I catch myself doing the same thing. Yet, since you strive for scientific accuracy in your blog, and deal with many topics that boarder on the psychological literature, I thought I'd chime in. :)

Stephan Guyenet said...

Thanks Aaron-- good stuff. It make sense but I'm going to have to let that percolate for a while.

2lbsofstarch.com said...

I would occasionally binge on the Double-Stuff Oreos (I wouldn't waste my time on the original ones), but I generally preferred the savory stuff (e.g., a bucket of Extra Crispy). What's the evolutionary basis for men to crave savory and women to crave sweets? Also, if you showed a much higher reward in rats for sugar (carbs) vs. fat, would Gary Taubes be happy?

myfiredreams said...

Very glad to read this- and I feel perfectly comfortably calling it an addiction, because I used to be addicted to sugar. I have to stay away from it completely- that was my only way to get over it. I used to entire 1/2 gallons of ice-cream in one sitting, or half a pan of brownies. I'm glad they are doing research on this- I think it's quite serious. I now maintain 18% bodyfat as a female and have been crossfitting for 2 years- but I could never have made this change if I had not cut off sugar completely- trying to keep it in in "normal" amounts was just too difficult- like your article said- it's an addiction because your need for it supercedes the importance of your family and responsbilities in life. It was a bad time in my life when I was addicted to sugar. Anyone who truly can't eat it in moderation and knows, please think about getting off it cold turkey- I am craving free now by not consuming it. As soon as I start having a little again, I instantlly want to eat whole boxes/pans/cartons of the stuff again. It's actually so much easier to not eat the junk at all. Now I actually crave raw kale salads- really no joke! Glad to find your blog.