Eating, Stress, Reward, and Obesity

Feb 14 2011 Published by under Behavioral Neuro

Sci recently covered a mouse paper on dieting and subsequent high fat eating in mice, and then she found THIS paper, on stress reactions in overweight humans! And I like how the two link up. So let’s take a look at this one, and then go back to the OTHER one, and see how they fit together.

We are told all the time that obesity is a public health concern, and we’re often told that causes include close proximity to McDonald’s, and the fact that no one plays outdoors anymore. But there’s also a correlation between obesity and stress. High levels of stress are associated with increased body weight, and in times of stress, we tend to go for high fat foods (aka comfort foods).

Comfort foods are thought to ACTUALLY comfort you and relieve your stress, by reducing stress hormone levels like those of corticotropin releasing factor. And stress is also an indicator for FUTURE weight gain, the more you’re stressed, the more you’re probably GOING to be stressed, and the more weight you may end up gaining (grad students around the world are DOOMED).

So how can we change this? Well, one way is by changing the food, but diets don’t work, and may wonk out your stress system. What we COULD do is change the way we respond to stress.

And that might start with changing the reaction of our ventral striatum.

Jastreboff et al. “Body mass index, metabolic factors, and striatal activation during stressful and neutral-relaxing stated: an fMRI study” Neuropsychopharmacology, 2011.

So we know that in times of stress you’re likely to eat. The eating makes the stress feel better, in part by hitting stress pathways, but ALSO by hitting reward-related pathways. In particular food (and sex, and drugs, and now rock and roll) activates the areas of reward circuitry involving dopamine in your brain. Dopamine is a neurotransmitter that is involved in pleasurable responses to stimuli like food (it’s involved in lots of other things to, but this is what we’re focusing on for the moment). Dopamine neurotransmission is also screwed up in states like drug addiction, as well as states like obesity. In fact, changes in dopamine receptors in areas like the ventral striatum (a big area in reward and dopamine) happen in people with binge eating and obesity in much the same was as happens to people with drug abuse problems.

So you’ve got the stress dysregulation, you’ve got dopamine dysregulation. How do the two interact?

To look at this, the scientists in this study look a bunch of normal weigh and overweight people, and put them in an fMRI scanner, which looks at blood flow to various brain areas in real time. They wanted to look at the response of the ventral striatum in response to stress in these people, as well as responses to normal states and relaxed states. So they picked three things, and presented them to the patients. They had a normal state, where the patients were read cued scenarios like remembering a day at the beach or walking in the park. They had a stressed state (previously submitted by the patients themselves) where they were told about a breakup, losing their job, unemployment, or the death of a loved one. And they had a “relaxed” state, which in this case used an alcohol cue of meeting a friend for drinks or having a party (Sci is not thrilled with this cue, but I’ll tell you about it later).

And then they looked at their ventral striatums.

What you can see in the table above is the anxiety ratings that the patients gave for all of the scenarios. The overweight patients rated the high anxiety conditions as being significantly worse than the normal weight patients (I’d really like to know if they all got the same cues here, or if they used very personalized ones, because I think comparing stress in public speaking to the death of a loved one would be no comparison here…). The overweight individuals also showed higher anxiety responses to NEUTRAL conditions, which suggests that maybe they are more anxious as a group, which they might want to pursue more cause I think that’s a REALLY interesting finding.

Now to the MRI.

The results in the MRI were similar to the self-rating of anxiety that the patients gave. The overweight participants showed higher activation in the ventral striatum in the normal and in the stress conditions. This finding is a little odd to me, as I’d expect the activation to not be related to the stress itself, but activation in the ventral striatum to be more related to things like craving, which could be induced by stress.

This was in fact what they were after, the authors state in the discussion that the increased activation of the ventral striatum might influence the desire for rewarding things like food.

So it appears that people who are overweight have increased ventral striatum activity in response to stress. Questions remain about things like whether they would have this response if they weren’t overweight, and which came first, the weight or the anxiety response.

I’ve got three issues with this study.

1) I feel like they didn’t go as far as they could have. Sure, they have different responses to stressful and neutral states between overweight and normal weight individuals, but could they change it? What if they trained the people in relaxation techniques or gave them therapy? Would the response to stressful states decrease in both the overweight and normal weight people? What about neutral states?

2) I really didn’t like the alcohol cue, and I didn’t particularly enjoy the neutral cue either. They did deliberately pick people who were “social” drinkers and consumed less than 7 drinks a week, but still. First off, not all people find drinking relaxing. Secondly, alcohol itself is a drug and a substance with significant social weight, and whether or not people are alcoholics, it has more significance than just relaxation. Finally, just because people SAY they drink less than 7 drinks a week, doesn’t mean they actually DO, and you can drink less than 7 drinks a week and still have an issue (if a girl drinks all seven of those in one night per week, for example). I feel like this study might have been better using different cues, perhaps provided by the participants. Relaxing things like puppies and kittens and playing with your kids…and reading a novel in the park. Cause that’s the other thing, I think that their neutral states might have been too relaxing. For many people, going for a walk or reading a book are ways to relax and might have that association. I feel like a more neutral condition might be a cue like, say, preparing dinner, filling out boring paperwork, or waiting at the hairdresser or something. I think then they might have had more interesting differences, but I’m not sure.

3) I’m not sure why they are looking in the ventral striatum and trying to correlate it with measures of anxiety. The ventral striatum and anxiety in generally don’t tend to go together much (you’d think they’d be wanting to look in the amygdala, which is associated with a lot of fear and anxiety associated responses). And sure enough, they didn’t get too much of a response (the differences are pretty small). I know that they want to relate it to things like overweight people having higher anxiety and then the activation in the ventral striatum driving them to seek out food to make them feel less stress, but if this is the case, why didn’t they ask the participants what they wanted to EAT?! Like how hungry were they, how much could they go for a pizza right now, do they CRAVE food? Especially after exposure to the stressful condition, the results for this, correlated with activation in the ventral striatum, might have been really interesting.

But now let’s get back to that mouse study previously, showing that stress causes mice to overeat, and that diet makes it worse. Those mice also showed an increased corticosterone response to stress. Unfortunately they didn’t do any anxiety testing on them, but it is nice how these two studies go together. In mice AND humans, stress drives us to seek a high fat diet. I wonder if the methylation changes in the mice which changed their food responses also exist in the humans? The behavior of the humans seems to indicate that might be a possibility. Remember though, the mice had been on diets and then exposed to high fat, but it’s very possible that the overweight people used in this study had a history of dieting. It’d be very interesting to see if the DIETING had anything to do with it, and how overweight people who have previously dieted differ from overweight people who haven’t gone on a diet. Not only that, it’d be interesting to see how PEOPLE eat differently when exposed to stress, depending on weight and whether or not they’ve been dieting. While we can’t look at the DNA methylation in the brains of humans (people tend to object to that), the mouse data may well tell us what could be happening to the humans responding in studies like this, and may tell us how to help change their responses.

Jastreboff, A., Potenza, M., Lacadie, C., Hong, K., Sherwin, R., & Sinha, R. (2010). Body Mass Index, Metabolic Factors, and Striatal Activation During Stressful and Neutral-Relaxing States: An fMRI Study Neuropsychopharmacology, 36 (3), 627-637 DOI: 10.1038/npp.2010.194

7 responses so far

  • Pharm Sci Grad says:

    Grad students are DOOMED. Thanks for that. ;) One less thing to stress about if stress is going to make it worse. I'm trying running, hopefully that is more helpful than dieting, cuz this dissertation stuff kind of sucks. LOL

  • Rich and Co. says:

    Here are some items"

    - We are contra the dopamine = reward model and pro (inherited impaired) dopamine receptors = hyper-seeking frame. Subtle but we think important difference.

    - What comes first? We are not big on environmental factors as causes rather than effects. For example, there is research that comfort food especially soothes (reduces hyper-seeking) of people abused as children. So the stress potentialities is likely set at birth and in early childhood to play out the rest of ones' life.

    That research is from AU and we can search for it -- not widely reported.

  • FrauTech says:

    Interesting study and thanks for going through it. I agree on the alcohol thing. I had to take a drug awareness test at work and it asked whether certain habits made someone an alcoholic. The test decided that drinking a single six pack of beer in one night did and drinking a glass of wine with dinner every night didn't. But in my mind, if you go to a party once a year and drink a six pack (depending on your weight and size) that may not be a drinking problem versus "needing" a glass of wine with dinner every night...so even the people trying to set the guidelines have their own ideas.

    • scicurious says:

      YES. This is something that gets me a lot. "Moderate" drinking means different things to different people. I know many alcohol reasearchers are not satisfied with the current guidelines and are looking at research that may change them, particularly with regard to women. Most alcohol guidelines for women right now are "downsized" from those for men, just take the amount a man can safely drinking and adjust for lower weight, but that's clearly not an accurate assessment in many cases.

  • can you at least says:

    can you at least pay a few bucks for the image? someone worked on that...

  • […] induce responses in the public through advertisements, religion, and media coverage. I also believe this kind of conditioning is one of the reasons why we have such a high rate of obesity in the world. For many generations, children were given ‘comfort food’ when they felt sad or […]

Leave a Reply


− eight = 0