We often think of anorexia as a psychiatric problem, a problem of self esteem, a problem of disordered body image. And while it's probably a lot of these things, treatments based on body image improvement and self-esteem can only do so much. When you look for new treatments, where should you look? What systems in the brain should you look at to try and understand anorexia?

Could anorexia be a disorder of reward?

Frank et al. "Anorexia Nervosa and Obesity are Associated with

Opposite Brain Reward Response" Neuropsychopharmacology, 2012.

Obviously, it's much more complicated than just a disorder of reward. But when people who are anorexic fall into disordered eating patterns, rewards themselves may become different. That was what the scientists wanted to look at in this paper.

They took 21 people diagnosed with restricting type anorexia (which is different from binging type), 19 obese individuals, and 23 controls. They wanted to look at reward responses in these patients using fMRI.

First, they tested them on taste. Sucrose solution vs "artificial saliva" (ew). No difference in the perception of pleasantness of the sweetness. Then they put them in the fMRI.

The fMRI test was actually a behavioral test of prediction errors, all the patients had to do was watch a screen and receive tastes of sucrose solution or artificial saliva (again, ew, why not water?!) in their mouths. The sucrose usually appeared after a particular shape on the screen (a conditioned stimulus), but sometimes it didn't. The conditioned stimulus would cause the patients to expect a reward, and the brain should respond one way, but then during the error they would not receive one, and the brain will respond differently (for more on this, see my older post on your brain on rickroll). The authors of the study were looking for these positive responses to reward, and the responses to errors.

They found activity in the usual places associated with expectation of reward: striatum, insula, orbitofrontal cortex, and amygdala. Activity was associated with reward prediction (whether they were going to get sucrose solution), in all groups. But while it was associated in the controls and anorexics, it was LESS associated in obese individuals, suggesting a blunted effect.

In particular, the left orbitofrontal cortex showed reward prediction responses. And here they saw not only that obese individuals had less response in this area, anorexics had more. And during error prediction, the anorexic groups also showed more activation in the orbitofrontal cortex than obese individuals or controls.

So the anorexic group and the obese group showed oppositional responses in this reward prediction test. The obese group had blunted responses to reward (similar to that seen in other studies), but no difference in error responses, while anorexics showed a larger error response, and a larger reward response in the orbitofrontal. The authors then compared this to a model dopamine network, and concluded that anorexics and obese inidividuals have opposite dopamine responses in this reward paradigm.

I think it's a pretty interesting finding (and pretty much as predicted, if people who overeat have blunted responses, then shouldn't people who undereat have sensitized responses?). And if it does turn out to be an issue of reward networks, and particularly of dopamine, this could make a target for drug interventions for these patients. The authors also suggest using scans like this to detect anorexia, but I'm not sure that's as necessary as new treatment options.

But of course I've got a couple of issues:

. Why do they always use OBESE individuals? I think this is an issue in this particular experimental area. If you are trying to show the opposite of anorexia, trying to show binge eating...recruit BINGE EATERS. Not people who are obese. Obesity does not always mean binge eater, and binge eater doesn't always mean obese. I imagine there is some overlap, but it's clearly not 100%. It seems like a easy variable to control for. I think this is a big oversight in the field (though some studies do control for this). Though I do imagine it's easier for study recruitment to get people who are obese rather than people who are explicitly binge eaters.

2. Would have been nice to see PET here, and now just fMRI. PET studies, positron emission tomography, can look at the binding of a radiolabeled ligand to a specific receptor, giving you an idea of system function rather than just 'this area lights up more or less'. PET does not have the excellent anatomical specificity of fMRI, but the two together are a powerful combination. If they had the addition of PET, they could have looked at dopamine in particular, which in reward-related areas, would have given the study extra oomph.

3. They stuck with taste only. Obviously, since this is a study on eating disorders, you can see why, but I'm curious, is this limited to food related rewards? What about, say, playing little games for small amounts of money? Does it always vary in the same way? I'd be very curious to see if the reward changes are limited.

But overall, I am glad to see this study. I've been seeing a lot of studies lately on reward-related systems and binge eating, and it's nice to see it "flipped around" for anorexia, to get at the other side of the behavioral coin (though obviously, neither disorder is that simple). And the more we know about potential mechanisms, the more steps we can take toward effective treatment.

Frank GK, Reynolds JR, Shott ME, Jappe L, Yang TT, Tregellas JR, & O'Reilly RC (2012). Anorexia Nervosa and Obesity are Associated with Opposite Brain Reward Response. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology, 37 (9), 2031-46 PMID: 22549118